Selective Incapacitation and Public Policy
SUNY series in New Directions in Crime and Justice Studies
Austin T. Turk...
33 downloads
748 Views
3MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Selective Incapacitation and Public Policy
SUNY series in New Directions in Crime and Justice Studies
Austin T. Turk, editor
Selective Incapacitation and Public Policy Evaluating California’s Imprisonment Crisis
Kathleen Auerhahn
State University of New York Press
Published by State University of New York Press, Albany © 2003 State University of New York All rights reserved Printed in the United States of America No part of this book may be used or reproduced in any manner whatsoever without written permission. No part of this book may be stored in a retrieval system or transmitted in any form or by any means including electronic, electrostatic, magnetic tape, mechanical, photocopying, recording, or otherwise without the prior permission in writing of the publisher. For information, address State University of New York Press, 90 State Street, Suite 700, Albany, NY 12207 Production by Michael Haggett Marketing by Jennifer Giovani
Library of Congress Cataloging-in-Publication Data Auerhahn, Kathleen, 1970– Selective incapacitation and public policy : evaluating California’s imprisonment crisis / Kathleen Auerhahn. p. cm. — (SUNY series in new directions in crime and justice studies) Includes bibliographical references and index. ISBN 0-7914-5797-4 (alk. paper) — ISBN 0-7914-5798-2 (pbk. : alk. paper) 1. Imprisonment—Government policy—California. 2. Criminal Justice, Administration of—California. I. Title. II. Series. HV9305.C2A84 2003 364.6'09794—dc22
10
9
8
7
6
5
4
2003057268
3
2
1
To Robert A. Hanneman, for his astounding intellect and generosity of spirit
This page intentionally left blank.
Contents
Acknowledgments Part I Chapter 1 Chapter 2 Chapter 3 Part II Chapter 4 Chapter 5 Chapter 6 Part III Chapter 7
ix
The Criminal in Society: Penal Response and Rationale Introduction 3 Criminal Punishment in Civil Society: Purpose and Method 17 Criminal Sentencing Reform and Paradigm Change in California 39 Incapacitation and Dangerousness Selective Incapacitation Dangerousness Assessing the Level of Dangerousness in the Criminal Justice System
61 73 87
Chapter 9
Evaluating the Past, Choosing the Future Modeling the California Criminal Justice System, Part I: Reproducing and Evaluating the Past Modeling the California Criminal Justice System, Part II: Predictive Evaluation Conclusion: Choosing California’s Future
131 153
Technical Appendix A
Data Sources and Estimation Procedures
159
Technical Appendix B
Example of Simulation Model Code
167
Chapter 8
103
Notes
171
Bibliography
187
Subject Index
215
Author Index
221 vii
This page intentionally left blank.
Acknowledgments
A project of this magnitude never represents only the efforts of one person. Along the way, many people helped me in various ways, some of which are evident in these pages, and others, while not necessarily apparent, were nonetheless instrumental, if only through their contribution in the effort to preserve my sanity. I would like to thank the following individuals. First, I am indebted to Bob Hanneman, for his support throughout the years. He is without question the finest teacher I have ever known. He’s probably also the smartest person I will ever encounter. Words cannot express the debt I owe to Bob for shaping me as a scholar and a teacher. Similarly, Austin T. Turk has taught, guided, and challenged me to be a better theorist and scholar. A seemingly innocuous notation on the margin of a manuscript would sometimes send me reeling off into weeks of research—which would manifest itself as a single sentence or note in the manuscript, but one that ultimately became essential to the argument I was making at the time. He’s been doing this to graduate students for years. I hope the others appreciate and grow from it as much as I have. I am grateful to Bob Figlio for his input in the early stages of this project. Thanks also to Shaun Bowler, who I managed to trick into giving me a lot of help and guidance during later stages. Much of the work on this book was completed while I was a graduate student in residence at the sociology department at the University of California, Riverside. I would like to thank the department staff for their assistance over the years: Renee DeGuire, Terry DeAnda, and Cathy Carlson. Anna Wire deserves extra thanks, not just for the myriad ways in which she assists the graduate students, but for being such a wonderful and supportive friend. I am also especially grateful to Robin Whittington for her extensive assistance in securing and administering funding to support this research. Validation data were provided by the California Criminal Justice Statistics Center (CJSC) and by the California Department of Corrections. Linda Nance
ix
x
Acknowledgments
of the CJSC was especially helpful. I am also grateful to Helen Ross and Alexis Alvarez, who both provided invaluable last-minute assistance with data entry. Alex Muentz deserves special thanks for being extraordinarily supportive and understanding of the long hours and occasional fits of short temper that preparing a manuscript for production entails, and for loving me in spite of these things. I was fortunate to receive the generous support of the National Institute of Justice for this research. This project was supported by grant number 1999-IJCX-0043, awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. The points of view in this document are those of the author and do not necessarily represent the official position or policies of the U.S. Department of Justice.
Part I The Criminal in Society Penal Response and Rationale
This page intentionally left blank.
Chapter 1 Introduction
This work is about criminal punishment. It is about the philosophical rationales that underlay this practice, and the way in which these justifications shape the methods employed in the enterprise of punishing criminal offenders. It is also about the outcomes that result when the theoretical purposes and operational realities of a system of criminal punishment meet blindly, proceeding without any explicit reference to one another. Many have argued that the current state of the entire enterprise of criminal punishment in America is a textbook case of “unintended consequences.” These days, prisons are generally perceived as institutions that do not appear to do much, if anything, about the level of crime in our society; nor are prisons widely hailed as institutions that are capable of effecting positive changes in the behaviors or attitudes of the offenders housed within them. Yet, these institutions are currently so central to the system of criminal sanctioning that a majority of state correctional systems contain populations that far exceed capacity limits. Surely these circumstances could not have been intended by anyone. This work attempts to explain how and why this situation came to pass, and to evaluate the system’s performance in terms of its own explicit objectives. That the American criminal justice system produces unintended and perhaps undesirable consequences is not a novel observation. In fact, a fair amount of attention has been devoted to this very idea. What is perhaps most striking about this body of research, taken as a whole, is the overwhelming lack of recognition of the systemic character of the criminal justice system, and the effects of this systemic quality on the translation of reforms into operational policies. Important exceptions do exist. For example, Feeley and Simon (1992) have offered an explanation of the ways in which correctional authorities have 3
4
The Criminal in Society
attempted to manage and adapt to the unintended consequences of sentencing policies in the process of carrying out the daily burden of custody by refocusing the conceptual orientation of the sanctioning process at the operational level. Similarly, Hepburn and Goodstein (1986) and Bales and Dees (1992) have examined how the intent of legislative sentencing reform is often distorted by the realities of implementation, much like a child’s game of “Telephone,” in which a phrase is whispered successively into the ears of a line of children, and frequently emerges at the other end as something that bears little resemblance to the original utterance. This work focuses on a particular form of criminal punishment, imprisonment. Imprisonment arguably occupies a central position in criminal justice in the hearts and minds of most members of society (Dershowitz 1976; Foucault 1977). Over the past three centuries in American history, the justification of imprisonment as a criminal sanction has been defined by the goals that the incarceration of offenders is expected to achieve. These goals become institutionalized into widely accepted paradigms, which define an era in the operation of the criminal justice system (Kuhn 1996).1 Paradigms of criminal punishment are founded upon assumptions about the causes of crime; these assumptions are in turn based on conceptions of human nature—and, by extension, the nature of criminal offenders (Wilson 1983). The understanding of the criminal act and the criminal offender quite naturally guides the selection of criminal justice responses to crime. Criminal justice paradigms are most prominently expressed in criminal sentencing policy (Hewitt and Clear 1983:24). For this reason, my analysis focuses on criminal sentencing reform as the clearest expression of paradigm change. Often, reforms are accompanied by explicit statements affirming the new paradigm and/or rejecting the old; an example of this can be seen in the text of the Comprehensive Crime Control Act of 1986, in which Congress officially announced its disdain for “the outmoded nineteenth-century rehabilitative theory that has proved to be so faulty that it is no longer followed by the criminal justice system” (Congressional Information Service 1986). Ultimately, I argue that incapacitation has emerged as the principal objective of criminal sentencing policy today. This book is not a philosophical undertaking; after exploring the historical trajectory that has established incapacitation as the dominant paradigm in criminal justice, the moral and ethical merits of incapacitation as the rationale for a system of criminal punishment will not be debated.2 The purpose of this study is primarily an empirical one. Having demonstrated the prominence of the incapacitation paradigm in contemporary criminal justice (I argue that the incapacitation objective manifests itself in recent sentencing reforms in a highly specific form, selective incapacitation), the task is then to evaluate the operation of a sanctioning system in terms of this objective. For this purpose, the analysis focuses on the criminal justice system of the state of California.
Introduction
5
California was chosen for the analysis for several reasons. With over half a million adults under some form of correctional supervision, California has the largest criminal justice system in the United States (Maguire and Pastore 2001). Although approximately 8% of the total U.S. population resides in California, the state’s correctional facilities house nearly 15% of all prisoners in American state and federal institutions (Gilliard and Beck 1998). Currently enumerating over one hundred sixty thousand inmates, California’s prison population has more than quadrupled since 1980 (California Department of Justice 2002). It has been asserted that the majority of this increase has resulted from changes in criminal justice policy, rather than from changes in crime rates, which have remained relatively stable over the same period (Irwin and Austin 1994; Zimring and Hawkins 1994). California’s criminal justice policy arena is a particularly volatile one. Amid the flurry of habitual offender statutes that has swept the nation in recent years, nearly 72% of California voters passed the most broadly written and widely implemented “Three Strikes” law despite ballot disclaimers stating that the impacts on crime, as well as the fiscal consequences of such a policy were “unknown.” An examination of the consequences of criminal justice policy reform in California is also worthwhile if one puts any stock in the idea that the state plays an agenda-setting role in the national public policy arena (Foster 1996). If, for example, there are states contemplating California’s approach to Three Strikes, then an evaluation of the consequences of this approach might be helpful to other legislators formulating their own policy choices. Finally, an additional reason for focusing on a single state is largely a methodological one. It is my belief that analyses of criminal justice policy are best conducted at this level of aggregation. It is a misnomer to speak of “the American criminal justice system,” when in reality the “system” is comprised of fifty-one independent systems (i.e., the states and the federal system). This study seeks to fill in some of the gaps in the existing literature on the efficacy and effects of sentencing reform. In the past decade or so, a great deal of theoretical and empirical research has taken place in the area of criminal sentencing. Some of these studies have examined the success of sentencing reform with respect to the implementation of reforms (Wichayara 1995; Ulmer 1997; Austin et al. 1999). Some researchers have looked at the question of crime reduction impacts resulting from get-tough sentencing policies (Clear 1994; Spelman 1994; Wichayara 1995; Zimring and Hawkins 1995; Stolzenberg and D’Alessio 1997; Zimring, Hawkins, and Kamin 2000), while others paint with a broader brush, and analyze sentencing policy reform from a cost/benefit perspective (Zedlewski 1987; McIntyre and Riker 1993; Baum and Bedrick 1994; Greenwood et al. 1994; Connolly et al. 1996). Another prominent area of study is the impact of sentencing reform on racial disparity in the criminal justice system (Schiraldi and Godfrey 1994; Tonry 1995; Davis et al. 1996). In addition
6
The Criminal in Society
to these empirical studies, a sizable literature has developed in the past decade that evaluates the value of existing policy goals from a normative or theoretical standpoint (e.g., Walker 1991; von Hirsch and Ashworth 1992; von Hirsch 1993; Clear 1994; Palmer 1994). While the present work is informed by all of these contributions, it explores a territory that is rather different from that which has been investigated in previous works. Rather than considering the normative propriety of selective incapacitation as the primary goal of criminal punishment, I consider the prominence of selective incapacitation in penal purpose as a “social fact,” and examine the efficacy of criminal sentencing policy in terms of this objective. Two chapters of the book are devoted to offering an explanation of how and why selective incapacitation has come to supplant other goals of criminal punishment in the American consciousness. The primary empirical objective of the study is to evaluate sentencing policy in California with respect to this objective. Recognizing the limitations of the methods used previously in the literature, which include the estimation of crime-rate reductions and the calculation of “social costs and benefits,” I develop an evaluation strategy that focuses on the selective success of incapacitation policies with respect to dangerous offenders. This approach might be characterized as an exercise in “putting California’s money where its mouth is.” Simply put, I will seek to discover whether or not sentencing policy reforms that aim to protect the public by incapacitating dangerous offenders, such as Three Strikes and You’re Out and Truth in Sentencing, have indeed been successful in incarcerating such offenders.3 Although a sizable portion of these analyses focuses on the history and consequences of California’s Three Strikes statute, this work is not a study of Three Strikes per se—rather, Three Strikes is considered as merely a prominent example of the types of laws that emerge in the context of larger trends in sentencing policy. The first substantive chapter of the book (chapter 2) examines the ideological and operational trends in the history of criminal punishment. While this discussion focuses primarily on the United States, attention is directed abroad when other nations influence and house the origins of American practices. This chapter also provides an account of twentieth-century trends in criminal justice leading up to the most recent penal paradigm, selective incapacitation. Michael Sherman and Gordon Hawkins (1981) note that thinking about policy choices that is engendered by the “crisis mentality” and by the search for quick-fix solutions is often ahistorical in nature. They remind us that It must be remembered that correctional populations result from decisions based on qualitative, normative assumptions. The prison population rises not by some mysterious levitation but because society, through its agents, decides that certain people ought to be locked up. To see the prison crisis exclusively as a problem of crowding and conditions is pos-
Introduction
7
itively dangerous. It addresses effects while ignoring causes. (Sherman and Hawkins 1981:4; see also Zimring and Hawkins 1991, chapter 3) A number of scholars have identified a growing emphasis on actuarial strategies of risk reduction and reallocation in crime control (Reichman 1986; Simon 1987, 1993; Feeley and Simon 1992; O’Malley 1992), as well as in the larger society (Beck 1992; Douglas 1992). The prominence of selective incapacitation in contemporary penal purpose is a logical expression of these trends. In the words of Malcolm Feeley and Jonathan Simon, “incapacitation promises to reduce the effects of crime in society not by altering either offender or social context, but by rearranging the distribution of offenders in society” in such a way that probabilities and risks are altered in the general population (Feeley and Simon, 1992:458). The present research differs from previous analyses of “risk society” in that most prior work in this area has tended to focus on abstract, “Foucauldian” (O’Malley 1998) conceptions of risk; these analyses emphasize the social meaning of risk and the societal responses to its occurrence. This research focuses more concretely on the notion of dangerousness as it applies to criminal offenders and penal responses to crime. Chapter 3 addresses itself to demonstrating how the process of philosophical and ideological evolution outlined in the previous chapter has manifested itself in California sentencing law. Particular focus is given to the legislative reforms of the last half of the twentieth century. There are two reasons for this. First these reforms are responsible for the enormous changes in the sheer magnitude of prison populations. The second reason is simply that in California, as in the nation as a whole, little novelty was in evidence in the “science of penology” for nearly two centuries in terms of beliefs about the best way to deal with criminals. 4 It was not until the early 1970s and the onset of the nationwide phenomena that Francis Allen has termed “the decline of the rehabilitative ideal” that states undertook sweeping programs of reform in their criminal justice systems (Blumstein et al. 1986). The idea of incapacitating criminals is hardly a new one; indeed, as Morris and Rothman (1995) suggest, since the “system of trials presupposes the existence of a jail [to secure the accused’s appearance] . . . the original justification for the prison may well have been incapacitation” (Morris and Rothman 1995: ix). Chapter 4 delves a bit more deeply into the idea of incapacitation, and argues that due to the apparent infeasibility of a strategy of collective incapacitation (demonstrated in a number of widely publicized studies), it is selective incapacitation that has captured the imagination of policymakers and their constituents. This is apparent in the focus on “career criminals,” “habitual offenders,” and “violent predators” that pervades the public discourse about crime and criminal justice today. Chapter 4 details the emergence of selective incapacitation in the research and policy arena. Attention is also devoted to some of the
8
The Criminal in Society
legal policy prescriptions deriving from this idea, with particular focus on the California experience. The ultimate goal of selective incapacitation is the reduction of crime. The strategy traditionally employed to evaluate the effectiveness of selective incapacitation focus on crime rate reductions attributable to incapacitation-oriented sentencing policies. Crime rate reductions are usually calculated as a summary function of the average rate at which high-rate or dangerous offenders commit criminal offenses (), multiplied by the number of such individuals assumed to be incarcerated under a policy of selective incapacitation. The crime-reduction impacts of selective incapacitation are ostensibly accomplished via the incarceration of “high-rate offenders,” “career criminals,” or some other name given to a class of offenders who are believed to contribute disproportionately to the total volume of crime. In chapter 5, I assert that the various names given to the targets of policies based on the idea of selective incapacitation (e.g., “habitual offenders,” or “career criminals”) are synonymous with a single underlying construct: the dangerous offender. The traditional evaluation strategy assumes the intervening step—a step that this work problematizes and investigates—namely, that dangerous offenders are successfully targeted under these sentencing policies, thus resulting in a reduction in crime. This assumption is problematic for several reasons. First, it is entirely possible that a policy of selective incapacitation could be quite successful in targeting dangerous offenders yet fail to accomplish a reduction in the crime rate. This is due in large part to the failure of this calculation strategy to account for other influences on crime rates, such as the replacement of offenders or the effects of criminal groups (Blumstein et al. 1978:65; Spelman 1994; Zimring and Hawkins 1995). A second, and even more serious problem is the inherent artificiality in the calculation strategy and the sensitivity of results to foundational assumptions. The artificiality of these mathematical approaches is particularly well-demonstrated by the lack of consensus concerning estimates of ; published estimates of this value range from 2 to 187 offenses yearly per offender (see Spelman 1994:71–80 for a comprehensive review). Chapter 5 explores the notion of “dangerousness” and the nature of the process by which something comes to be considered dangerous. Turning our attention to the problem at hand, historical and contemporary conceptions of the dangerous offender are reviewed, as are a number of attempts to prospectively identify and control such offenders. These studies lead to the unmistakable conclusion that the prospective identification of dangerous offenders remains, as Norval Morris (1974) so delicately phrased it, “quite beyond our present technical ability” (Morris 1974:62). In chapter 6, I propose an alternate strategy for evaluating the efficacy of sentencing reform in terms of the proximate goal of selective incapacitation— that is, the incarceration of dangerous offenders. This strategy includes a concep-
Introduction
9
tualization of “dangerousness” for use in the retrospective evaluation of criminal sentencing policies. Dangerousness is here conceived as a stochastic property of populations rather than as a property of individuals. The probabilistic nature of dangerousness renders nonsensical a statement like “Offender A is dangerous” or “Offender B is not dangerous.” A statement along the lines of “Offender A is more dangerous than Offender B,” is less problematic, but in and of itself, this information is not terribly useful from a policy evaluation standpoint. It is both logical and instructive to conduct an analysis that allows us to say that “based on the known correlates of dangerousness, Population X is likely to harbor a greater proportion of dangerous individuals within it than is Population Y.” Because of this, the dangerousness construct developed in chapter 6 is designed to be used to compare the relative dangerousness of criminal justice populations, and also to assess changes in the level of dangerousness within a particular population over time. It is the logic of statistics and not that of prediction that characterizes this approach. In an essay entitled “Some Statistical Questions in the Prediction of Dangerous Offending,” John B. Copas (1983) points out that The absurdity of expecting precise predictions of individual behavior has already been stressed. . . . Although the outcome of tossing a coin is quite unpredictable, everyone will agree that to start a sports contest by the toss of a coin is “fair”. This is because the chaos at the individual level is replaced by an order at the group level. (Copas 1983:136) What may seem like a graceful linguistic maneuver is really of crucial analytic importance. Prior attempts to measure dangerousness (most notably the 1982 Rand report Selective Incapacitation) have proceeded as if dangerousness were an absolute quality instead of a relative one. Since dangerousness is a subjectively defined characteristic, 5 it is most sensibly considered in relative terms. In other words, we cannot say with certainty that a given individual is dangerous, only that he or she is more or less likely to be dangerous than another. The same is true of populations. While we cannot say with any measurable degree of certainty that California’s prison population in 2003 contains even a single dangerous individual, the measurement strategy I offer allows us to comment with some confidence on the likelihood that this population contains a greater or fewer number of such individuals (relative to the total population size) than did the comparable population in 1980. In order to assess the impacts of sentencing reform in California with respect to the objective of selective incapacitation, dynamic systems modeling is employed as the primary investigative tool. Although criminologists and sociologists commonly refer to “the criminal justice system,” empirical research in criminology tends to take the form of static or time-series analyses of single
10
The Criminal in Society
components of the system (e.g., jails, prisons, and courts) rather than conceiving of the entire system as a system (important exceptions include Blumstein and Larson 1969; Cassidy and Turner 1978; Cassidy et al. 1981; Cassidy 1985; and Ohlin and Remington 1993) . However, legislative changes that are intended to affect one component of the system may result in unintended systemwide consequences. For example, the primary aim of California’s 1994 Three Strikes law is the incarceration of “habitual offenders” for lengthy prison terms. However, this law has had a dramatic impact on many other parts of the criminal justice system as well. For example, trial volume has greatly increased, as defendants facing a second or third strike become increasingly unwilling or unable to plead guilty to lesser charges in exchange for reduced sentencing recommendations (Legislative Analyst’s Office [LAO] 1995). An additional consequence of the law is being observed in the state’s jails, where great numbers of defendants charged under the law are held awaiting trial, reducing available space for sentenced offenders (Turner 1999). Analyzing the impacts of legislative changes to sentencing structures from a systems perspective can provide important and useful insights into the unintended outcomes that result from these changes. A similar approach to the analysis of criminal justice systems was pioneered by Alfred Blumstein and his colleagues in the late 1960s and 1970s (Blumstein and Larson 1969; Belkin et al. 1972; Cohen et al. 1973). The JUSSIM model developed by these researchers represented the criminal justice system as a series of stocks and flows representing, respectively, phases or states that could be occupied by offenders (from committing a crime to being incarcerated in a facility), and what the authors call “branching ratios,” or the percentage of offenders that transition from one state to another at any given time. These models were not purely simulated, in that data were used to validate and parameterize the simulations. The modeling strategy was revolutionary in that it attempted to account for “feedback” of offenders through the system due to recidivism, and thus to delineate between crimes committed by “virgin” offenders and those committed by recidivists (Blumstein and Larson 1969). Later modifications of the model ( JUSSIM II and JUSSIM III) also focused on modeling the “careers” of victims as well as offenders (Blumstein and Koch 1978). The JUSSIM model has also been modified by R. Gordon Cassidy (1985) as the CANJUS model, which is used to study the operation of the Canadian criminal justice system.6 While the models I develop are structurally quite similar to the JUSSIM and CANJUS models, the focus and logic of the analysis is rather different. The JUSSIM models were primarily concerned with understanding the process and determinants of criminal careers, and on the impacts of criminal activity on law enforcement workload. The analyses conducted by Cassidy using the CANJUS model are somewhat more similar to the present work, in that he focused on
Introduction
11
processes of adaptation in the face of system change (Cassidy 1985; see also Cassidy and Turner 1978). The approach to modeling the California criminal justice system I develop in the work that follows is explicitly unconcerned with the processes that generate populations of criminal offenders; for this reason, it is perhaps more appropriate to consider this work as an analysis of the criminal sanctioning system. A detailed understanding of process is not necessary to the central question explored in this work. Rather, my goal is to faithfully reproduce the emergent structures that arise out of these processes, making use of data to validate the analyses.7 The analysis of dangerousness in criminal justice populations thus relies on a census-like logic—it is the composition of the populations that is of the utmost importance; a deep understanding of the nuances of the process that create these populations, is, in a sense, epiphenomenal. The modeling strategy employed in the analysis uses Berkeley Madonna software to construct a simulation model of the sanctioning processes in the California criminal justice system. This approach is based on the systems dynamics approach of Jay W. Forrester (1969a, 1969b) as explicated in Hanneman (1988). Chapter 6 details the particulars of the methodology. The system is comprised of states and rates. The states are population-states occupied by individuals within the system. These include the arrested population, the jail population, and the prison population. Figure 1.1 shows a simplified schematic of the systemic model. The rates represent the probability of an offender moving from one state to another (e.g., moving from the state of being arrested to the state of being detained in jail). The transition rates in the system are potentially dynamic, in that the simulation methodology allows for the modeling of the effects of feedback and informational processes on these rates. The outcomes that result from the operation of the system, such as the size and composition of correctional populations, are thus the result of the movement of individuals through the various states comprising the system. However, all individuals are not alike with respect to their experience of the criminal justice process. For example, black male offenders are more likely to be detained prior to trial than are white female ones.8 It is important to recognize that while some of these differences will coincide with indicators of dangerousness, the determinants of differences in transition rates need not be conceptually related to offender dangerousness. Indeed, in this analysis, offender attributes that contribute to differences in movement through the system are overtly considered to be unrelated to offender dangerousness, despite the fact that some indicators may overlap (e.g., sex).9 The necessity of taking into account all of the factors that are relevant to dangerousness and those that influence transition probabilities results in 450 different subpopulations. This is one reason why simulation modeling is preferable to attempting to directly estimate the system dynamics with actual data. The equation system that corresponds to the path diagram represented in figure 1.1 must be simultaneously estimated for each of these 450 subpopulations.
12
The Criminal in Society FIGURE 1.1 Structural Model of the California Criminal Justice System
An equation system with such a high degree of complexity simply cannot be estimated using direct mathematical methods—at least not without making many simplifying assumptions that have only tenuous theoretical justification. A common way of circumventing this problem in applications of structural equation modeling involves the imposition of a number of simplifying assumptions. However, I believe that the condition of California’s troubled criminal justice system has resulted in large part from a failure of researchers, politicians, and practitioners to attempt to conceptualize the system in all of its complexity. The skeptical reader may claim that I am defending a fictional method (i.e., simulation modeling) by highlighting the fictional qualities of another. However, although the modeling strategy employed in this study is indeed “simulation” (and therefore bears some resemblance to fiction) the first stage of the modeling process consists of replicating the existing system with reference to validation data, and is therefore grounded in empirical reality. Chapter 7 reports the results of the model that re-creates the compositional dynamics of the California criminal justice system from 1979 to 1998.10 This application of simulation modeling differs from other, more purely theoretical applications of simulation modeling (e.g., Jacobsen and Bronson 1985; Hanneman 1995) in that the construction of the simulation model is conducted with
Introduction
13
explicit reference to actual criminal justice data. The purpose of this exercise is twofold. The first goal is simply to evaluate the California’s criminal justice system’s efficacy in incarcerating dangerous offenders. The dangerousness construct developed in chapter 6 allows for the comparison of dangerousness levels in California’s correctional populations (prison, jail, probation, and parole) before and after specific criminal justice reforms.11 The second reason for modeling the system as it exists today is to gain an understanding of the system dynamics that have produced particular (objectively verifiable) outcomes with respect to the composition of criminal justice populations. The existence of a working baseline model allows for the component of the study that I call predictive evaluation. Chapter 8 reports the results of experimental projection analyses designed to evaluate the potential effects of recent criminal justice reforms that have the explicit intention of incarcerating dangerous offenders, specifically the state’s Three Strikes law. These analyses differ from ordinary population projections in several important ways. Most population projections rely on simple linear extrapolation of existing trends; however, as Zimring and Hawkins (1994) have pointed out, there is a great danger in making simple population projections in volatile periods of system growth (see also Greenwood et al. 1994, 1998). My strategy of predictive evaluation, in taking into account the dynamics of the entire system, differs from linear population projections in that I am modeling not only changes in the absolute numbers of inmates under various forms of correctional supervision, but the composition of populations—with respect to both dangerousness and demographic characteristics—as well as the processes that give rise to these outcomes. The simulation methodology makes it possible to explore the consequences of a variety of different potential policy choices—while being explicit about the assumptions underlying those choices. To use the metaphor of Sherman and Hawkins (1981), simulation modeling makes it possible, instead of simply predicting the future, to choose the future. In addition to estimating the likely consequences of the continuation of current sentencing practices, the simulation modeling strategy also allows for experimentation on the system to investigate “possible futures”—specifically, ways in which California’s system dynamics might be altered via changes in sentencing policy and practice to more effectively utilize its limited incarceration resources to selectively incapacitate dangerous offenders. My ultimate aim in offering this analysis lies in the hope that it will refocus thinking about penal strategy in a direction that is based on an analytic and realist criminology. David A. Jones identifies the roots of the analytic tradition in criminology in the work of Emile Durkheim (1982), who observed in The Rules of Sociological Method: Imagine a community of saints in an exemplary and perfect monastery. In it crime as such will be unknown, but faults that appear venial to the
14
The Criminal in Society
ordinary person will arouse the same scandal in ordinary consciences. If therefore that community has the power to judge and punish, it will term such acts criminal and deal with them as such. (Durkheim 1982:100)12 Other prominent scholars in the analytic tradition in sociology and criminology include Sellin (1938), Merton (1938), Vold (1958) Dahrendorf (1958, 1959), Clinard and Quinney (1967), and Turk (1969; 1982). These authors emphasized the role of conflict, power, and privilege in formulating definitions of crime and responses to the offender. Austin T. Turk (1969) framed the problem of criminality as a process of normative definition that emerges out of a pattern of conflict between authorities and subjects: The legality of cultural norms thus depends on how they are defined by authorities: a cultural norm is a law if the authorities say that it is, meaning by this that they are prepared to use their power against, to sanction, those who by their actions deny its relevance as a guide for behavior. Of course, the notion that everyone, authorities included, is bound in his own behavior by such a norm is an ideal limited to certain legal traditions and philosophies. Many norms are applicable to only particular categories of people, who alone are expected to conform; others are merely expected to accept the existence of such norms and the right of the authorities to enforce them. (Turk 1969:38) The analytic tradition is conceptually quite compatible with the newer “realist” school of thought in criminology. This school has variously been called “radical realism,” “left realism” (Young 1986; Lea 1992; Matthews and Young 1992), and “progressive realism” (Currie 1992). Realist criminology emerged in response to the explanatory poverty of the Marxist and postmodern approaches, as well as the theoretical and policy failures of mainstream criminology (Young 1986; Braithwaite 1989). The goals of radical realist criminology include the creation of a more comprehensive theoretical framework which can uncover the enduring processes that produce these problems and to provide a more solid basis for designing interventions . . . [realist criminology] considers itself to be radical in the sense that it draws freely on a tradition of critical theorizing which aims to demystify and dereify social relations. . . . [I]t is a criminology that expresses a commitment to detailed empirical investigation, recognizes the objectivity of crime, faces up to the damaging and disorganizing effects of crime, and
Introduction
15
emphasizes the possibility and desirability of engaging in progressive reform (Matthews and Young 1992:4; see also Lea and Young 1984; Young 1986; Currie 1992; Lea 1992; Lowman 1992) My approach to the evaluation of sentencing policy in California fits into the radical realist project in a number of ways. For one, the conceptualization and measurement of dangerousness does not deny the social reality of legally sanctioned categories and definitions of crime, but rather takes these as a “point of departure” and problematizes “the issue of ‘seriousness’ and significance of different crimes” as advocated by Matthews and Young (1992:5). Similarly, a strong commitment to the integration of criminological theory and practice is fundamental to radical realist criminology (Young 1986, 1992; Matthews and Young 1992). The main empirical objective is to evaluate California’s sentencing practices in relation to their explicit policy objectives, recognizing the validity of “rational democratic input” (Young 1992:49; Lea 1992). Rather than debating the legitimacy of the stated goals of California sentencing policy, the approach taken here accepts as a social fact the will of voters and their representatives in prioritizing the social defense objective as the primary goal of criminal punishment. The results of the retrospective analyses reported in chapter 7 show that, from the standpoint of the selective incapacitation of dangerous offenders, the sentencing policies implemented in California over the last two decades have not been wildly successful. The average dangerousness of the prison population has actually declined since 1980, while the dangerousness of noncustodial populations has actually increased. These analyses also highlight the importance of looking at the effects of criminal sanctions from a systemic perspective. Forrester (1969b) observed that “. . . [i]ntuition is unreliable. It is worse than random because it is wrong more often than not when faced with the dynamics of complex systems” (Forrester 1969b:24). The results of the retrospective analyses are consistent with the results of other researchers (e.g., Bales and Dees 1992; Turner 1999), indicating that reforms, primarily intended to effect change in prison sentences and prison populations have far-reaching effects on other criminal justice system functions, such as jail, probation, and parole. The prospective analyses reported in chapter 8 are presented in an attempt to find policy solutions that might help the California criminal justice system better achieve the goal of selective incapacitation of dangerous offenders. These analyses indicate that the state’s 1994 Three Strikes law, touted as the get-tough measure that would make the streets safer once and for all, will actually do very poorly at fulfilling this promise. Other “possible futures” are also explored; these analyses reveal that simple modifications to the law, such as releasing elderly offenders prior to the completion of their minimum terms, and restriction of the
16
The Criminal in Society
“strike zone” to crimes of violence can improve the functioning of the system vis-à-vis the incapacitation of dangerous offenders. Finally, chapter 9 concludes the book with a discussion of the implications of the findings with respect to making criminal justice policy, and also some consideration of the way in which criminologists and sociologists ought to proceed if we want to forge a link between empirical research and public policy.
Chapter 2 Criminal Punishment in Civil Society Purpose and Method
The use of criminal punishment in Western societies has generally been justified as serving one of four purposes: rehabilitation, deterrence, retribution, and incapacitation. At varying points in U.S. history, each of these purposes have been dominant in the construction of the ideology that guides criminal justice policy. In the last century, there have been several major paradigm shifts in the prevailing ideology concerning the purpose of criminal punishment; as a result of these shifts, very different policy decisions have been made than those that might have been considered under a paradigm assuming a different primary goal of punishment. Although today imprisonment is the focal point of the system of criminal punishment in the United States and in other Western nations,1 this is a relatively recent development—and one that is directly related to the ideological evolution of penal purpose in the eighteenth and nineteenth centuries. This chapter will trace the origins and development of the prison as an institution of criminal punishment, as well as the penal paradigms and social justifications that underlie this development. Thomas S. Kuhn (1969) defines a paradigm as “a constellation of group commitments . . . to shared beliefs” about the nature of a particular phenomenon (Kuhn 1969:181–184). Although Kuhn’s analysis focused on paradigm change in the natural sciences, his analysis is equally relevant to a discussion of the evolution of penal ideology.2 In the case of penal ideologies, paradigms are characterized by shared beliefs about the goals and expectations of criminal punishment. Also implicit in these paradigms are beliefs about the nature of criminal offenders, and of the ways to reduce or prevent crime in society. Paradigms are broader than theories, in that paradigms determine the scope of
17
18
The Criminal in Society
theories that may emerge. Additionally, paradigms may unite diverse methods of inquiry (Kuhn 1996:43). In the case of the natural sciences, it is relatively easy to discern the boundaries of communities that are the sites of paradigm change. What is the community that experiences and shapes paradigms about crime, criminals, and punishment? It includes policymakers and “experts,” such as legislators and academics, but the reach of paradigms of criminal punishment is greater in scope. Since crime is a defined as a social problem, and is experienced and felt, at least on some level, by every member of society, the community relevant to the construction of paradigms in criminal justice includes every member of society. For this reason, popular sentiments and perceptions about the proper responses to criminal behavior are salient forces in paradigm change. According to Kuhn (1969), paradigm shifts occur when existing paradigms fail to provide satisfactory solutions to identified problems (Kuhn 1969:67–68). The idea of crisis precipitating a paradigm shift is demonstrated by the welldocumented “decline of the rehabilitative ideal” (Allen 1981). Rehabilitation had been established as the dominant paradigm in American criminal justice for the better part of two centuries. But after the infamous declaration of Martinson and his colleagues, after having conducted an exhaustive survey of rehabilitative programs and strategies, that “nothing works” (Martinson 1974; Lipton, Martinson, and Wilks 1975), the rehabilitative paradigm came to be discredited. As the prison seemed to continuously fail in its pursuit of the objectives of rehabilitation, the goals of incarceration as a criminal sanction changed shape and direction. These changes manifested themselves in a variety of sentencing reforms.3 The history of criminal punishment in Western societies has followed a pattern of development that can be explained in terms of the overarching paradigms concerning the place of punishment within society. Many scholars divide the most commonly invoked justifications (or goals) of imprisonment into two general categories, those that are utilitarian in nature and retribution or desert (Packer 1968; Wilson 1975; von Hirsch 1976, 1985; Moore et al. 1984; Mathiesen 1990; Walker 1991).4 Deterrence, rehabilitation, and incapacitation differ from retribution in that an important goal of criminal punishment under these paradigms is the betterment of society. These utilitarian (also called social defense) paradigms share the idea that punishment has an object other than the offender himself; criminal punishment is thus a means to an end. Another distinguishing feature of these utilitarian aims is that deterrence, rehabilitation, and incapacitation are all forward-looking, whereas retribution is rooted firmly in the past. In the retributive paradigm, the only information that is relevant in making punishment decisions is the criminal act committed by the offender. In this paradigm, punishment is an end unto itself.5 Although each of the commonly invoked justifications of criminal punishment have summoned forth a variety of specific strategies to achieve these
Criminal Punishment in Civil Society
19
stated objectives, penal ideologies can be said to have “natural parameters.” For example, penal systems that are characterized by the deterrence paradigm tend to focus on harshness as the most salient technical feature of punishment; for this reason, grotesque and brutal corporal punishments have historically been associated with the deterrence imperative. Similarly, because the retributive paradigm requires the suffering of the offender, the harshness of punishments is of central importance. On the other hand, imprisonment is most befitting of the objectives of both rehabilitation and incapacitation.6 This is particularly true in the case of rehabilitation, for which the prison was conceived as a necessary element in effecting positive change in the offender, right down to the details governing the prisoners’ daily routines via the mechanisms of institutional rules as well as the architecture of the prison itself (McGowen 1995). What follows is a brief historical sketch of the evolution of penal purpose in Western society. Although my ultimate concern is with penal ideology in the United States, we must first look to developments in England and Europe to understand the origins of American criminal justice. This account focuses primarily on the ideologies surrounding the use of the prison; however, the prison’s place in the enterprise of criminal punishment cannot be fully understood without reference to the penal practices that preceded it.
HISTORICAL ANTECEDENTS OF THE PRISON The Ancient Period The earliest recorded statements on the place of criminal punishment in society indicate that utilitarian considerations were foremost in justifying the punishment of offenders in ancient societies. J. Thorsten Sellin’s (1976) historical analysis of criminal punishment shows that for the ancients, the purpose of inflicting punishment was unequivocally forward-looking, rooted in the objectives of deterrence and rehabilitation. Sellin offers this excerpt from the writings of the philosopher Protagoras (481–411 B.C.E.): He who desires to inflict rational punishment does not retaliate for a past wrong which cannot be undone; he has regard for the future and is desirous that the man who is punished, and he who sees him punished may be deterred from doing wrong again. He punishes for the sake of prevention, thereby clearly implying that virtue is capable of being taught. (in Sellin 1976:13) Socrates (470–399 B.C.E.) offers a similar view: “the object of all punishment which is rightly inflicted should be either to improve and benefit its subject or
20
The Criminal in Society
else to make him an example to others, who will be deterred by the sight of his suffering and reform their own conduct” (in Sellin 1976:13). The criminal laws of ancient Rome reflect this deterrent emphasis as well. The Twelve Tables (451 B.C.E.) featured a host of brutal corporal punishments obviously intended to deter would-be offenders from engaging in proscribed conduct (Scott 1932; Sellin 1976). Early Roman law also incorporated the explicit notion of crime as an injury to the body social, rather than merely an affront to the immediate victim (Lobingier [1987] 1923). Accordingly, social defense objectives are emphasized in prescribed punishments. While the punishments for many property offenses are restitutive in nature, such as the statute that decrees that a person guilty of grazing livestock on another’s land is required to turn the animals over to the injured party “by way of reparation” (Law V; Scott 1932:70), many statutes specified corporal punishments with a more expiatory bent. An example is Law VIII, which declares that “When anyone publicly abuses another in a loud voice, or writes a poem for the purpose of insulting him, or rendering him infamous, he shall be beaten with a rod until he dies” (Scott 1932:70).7 Penal incarceration had virtually no place in the arsenal of criminal punishments available to the ancient authorities of Greece and Rome. While there are references to the prison (or to other places of confinement) in ancient writings, it is clear that these institutions were used to detain rather than punish. The prison was used to hold offenders awaiting trial, or as a place for convicts to await the execution of their sentences (Peters 1995).8
The Middle Ages The earliest written documents in the English language include the Kentish Laws of King Ethelbert, which enumerated crimes and punishments under seventh-century British law. These laws specified a broad range of monetary fines for most offenses, including murder (Hibbert 1963; Briggs et al. 1996). Hibbert (1963) argues that this particular punishment structure can be accounted for by the influence of the church, which had an interest in reducing the amount of violence and feuding between injured and injuring parties,9 and that also received a substantial portion of the fines. Hibbert’s analysis supports Sellin’s (1976) contention that all modern penal practices originated as punishments inflicted only on slaves, which later broadened in their applications to include servants, laborers, and other members of the lower classes. When offenders were unable to pay the fines prescribed in the criminal codes for a particular offense, the law provided for an alternate punishment. These punishments were almost universally corporal in nature (Hibbert 1963; Sellin 1976). Despite the existence of alternatives to pecuniary punishments, the primary function of punishment in the criminal justice system of early medieval England
Criminal Punishment in Civil Society
21
was the generation of revenue (Hibbert 1963; Briggs et al., 1996). It would remain so for several centuries. After the Norman conquest in 1066, the form and purpose of criminal punishment began to change.10 Crime came to be defined not as a matter between two individuals (a wrongdoer and a victim), but rather as a matter between the offender and the state. Criminal behavior was thus seen as above all injurious to the body social (Hibbert 1963). The principal result of the state’s appropriation of the victim role was an increase in the use of corporal punishments such as mutilation and execution (Hibbert 1963; Kuntz 1988). The frequency with which such punishments were applied continued to increase throughout the late medieval period (Hibbert 1963; Peters 1995; Langbein 1998). In addition to corporal punishments, penal slavery was a common sanction in the late Middle Ages. With the advent of sea warfare in the sixteenth century, enslavement on galley ships (“at the oars”) became a widely used form of punishment in Southern Europe (Sellin 1976; Langbein 1998). In large part the increase in corporal, and more importantly, public punishment was a reaction to “the people’s growing predisposition to crime” (Hibbert 1963:8), and as such represents a shift in the purpose of penal sanctions. In many ways, criminal punishment in the late Middle Ages represented a return to the deterrence paradigm exemplified in ancient Roman law (Peters 1995). Prisons also existed in one form or another in early medieval Europe. These institutions differed from the modern prison, in that punishment was not initially included in the legitimate purpose of such institutions; they served a custodial function and as such were auxiliary to the system of criminal punishment. Many scholars identify the origin of the modern prison in Europe in the institution of the workhouse,11 which housed a mixture of paupers, vagabonds, and petty thieves (McKelvey 1936; Kuntz 1988; Friedman 1993; Peters 1995; Rothman 1995; Langbein 1998). The state apparatus of criminal punishment grew all over Europe during this period in response to a virtual epidemic of vagrancy. This epidemic was the result of many factors, including the breakdown of the feudal system of farming, the growth of commerce, and increasing population and urbanization (Hibbert 1963; Peters 1995; Briggs et al. 1996; Langbein 1998).12
The Modern Era In early modern Europe, corporal punishments such as whipping (the most common), branding, mutilation, and execution continued to predominate. Throughout most European nations, corporal punishments, as well as other nonphysical types of punishment (which most often involved some form of ritual shaming) took place in full public view, for the edification and/or amusement of the townspeople (Foucault 1977; Spierenburg 1995). The deterrence paradigm guiding these brutal punishments was consistent with the so-called classical criminological thought of the eighteenth century.
22
The Criminal in Society
The classical school, epitomized by thinkers like Cesare Beccaria and Jeremy Bentham, held that crime was a product of individual choices, and that the function of punishment was to influence others from making similar choices (e.g., Beccaria (1983 [1775]). While the expiatory character of executions and other forms of punishment cannot be denied, the overwhelmingly public character of criminal punishment in early modern Europe attests to the prominence of deterrence as the primary objective of these punishments: Another element in the theater of punishment was the use of dead bodies as warnings to [the] living. Most European towns and villages kept a gallows field or gallows mountain on which they displayed the corpses of selected capital offenders. The bodies hung in public until they decomposed; those who had died on the wheel were propped up in the device, supported by a harness. Towns always located their gallows field at a conspicuous spot. (Spierenburg 1995:56–57; see also Hibbert 1963: chapter 2) Transportation was another common form of punishment in this period. Over the course of the seventeenth and eighteenth centuries, Great Britain transported approximately two hundred thousand convicts to the American and Australian colonies and to other colonial destinations such as the West Indies (Shaw 1966; Weidenhofer 1973). Transportation as a penal strategy initially developed as a response to overcrowded gaols housing minor offenders at home. The use of transportation to the colonies increased throughout the eighteenth century, after it began to be used as an alternative to the death sentence. Offenders convicted of certain offenses could elect to have their death sentences commuted to transportation to the colonies (Weidenhofer 1973). Transportation and penal slavery both represented a significant shift in the method of punishment away from corporal, public punishments. While slavery, confinement, and transportation entailed a considerable amount of suffering, this was, in large part, incidental to the actual sentence. Although many convicts were flogged, this punishment was generally received as a result of subsequent offenses committed either in the penal colonies or on the journey over, after the imposition of the sentence of transportation (Weidenhofer 1973). The structure, form, and purpose of criminal punishments in colonial America largely mirrored those in Europe during this period. Early American criminal law was influenced greatly by English criminal law, and as such was similarly focused on public forms of punishment such as shaming (e.g., the pillory and ducking) and corporal punishments such as whipping, mutilation, branding, and execution (Prince 1899; Chapin 1983; Friedman 1993; Rothman 1995).13 Although the colonies were the recipients of transported convicts (until the advent of the American Revolution), the colonists had an analogue in the
Criminal Punishment in Civil Society
23
simpler practice of banishment (Friedman 1993). Deterrence figured largely in justifications for criminal punishments in colonial America (DeWolf 1975). In addition to punishments like whipping and execution for serious offenses, public admonishment as a form of shaming was a popular sanction for more petty offenses (Chapin 1983).
Punishment and Modernity Michel Foucault (1977) has noted the increasing tendency for punishments to assume a private character in the modern period. In addition to noting the move away from public forms of punishment, Foucault also documents the shift in the nature of criminal punishment in the early modern period in terms of the object of punishment.14 Prior to the eighteenth century, corporal sentences of torture, mutilation, and execution predominated; in late modernity, penal strategies shift away from the body as the site of penal intervention. Modern punishments such as imprisonment, banishment, and forced labor involve the body, but actions on the body are a means to an end. The target of modern penal interventions is the mind or soul of the offender—in a sense, the offender’s humanity is the focus of attention, rather than his corporeality. Foucault notes that the method and purpose of criminal punishment changed in this period to a much greater degree than did the definitions of criminal offenses (Foucault 1977:17). Nathaniel F. Cantor (1932) asserts that the changing conceptions of human nature and natural rights that emerged during the Enlightenment also contributed to the decline in corporal punishments. However, this does not explain why, when support for the rehabilitative paradigm declines in the late twentieth century, floggings in the public square (or on its modern equivalent, pay-perview) do not return. As David Garland (1990) observes, “penal measures will only be considered at all if they conform to our conceptions of what is emotionally tolerable” (Garland 1990:214). Garland proposes that the changing relationship between sensibilities and punishment can be seen as part of a larger “civilizing process” in Western society (Elias 1982 [1939]); in this civilizing process, violence (and unrestrained emotional displays in general) becomes increasingly abhorrent. Norbert Elias (1982 [1939]) identifies the cause of this growth in “civility” in the increasing interdependence among individuals, which comes about as a result of the high degree of functional differentiation in modern society. Elias contends that sustaining these complex social arrangements requires greater constraints on the use of interpersonal violence. The other primary influence on the civilizing process is the monopolization of physical violence in the hands of a centralized state, which results in the creation of “pacified social spaces . . . normally free from acts of violence” (Elias 1982 [1939]:235). In so-called “civilized” societies, Elias argues, the taboos against
24
The Criminal in Society
violence are so strongly ingrained that they become part of the individual’s consciousness and personality structure.15 The key feature of “civilization” is the rationalization of emotion (Kuzmics 1988:155). Foucault’s account of the gradual replacement of brutal, public corporal punishments with other forms of punishment that are both more “humane” and increasingly hidden from public view is thus entirely consistent with the notion of the civilizing process. Nor does civilization preclude the return of the deterrence and retributive paradigms; there is a subtle but nevertheless important distinction between the experience of an emotion and the expression of one. The civilizing process does not create norms proscribing vengeance; it merely places regulatory limits upon the forms of its expression.16 Incorporating the notion of the “civilizing process” into the historical account of criminal punishment contextualizes the analysis of crime and punishment rather than isolating the penal system as an institutional structure apart from the rest of society. In so doing, I hope to improve upon the so-called revisionist formulations of the 1970s and 1980s, which challenged the hitherto prevailing narrative of the development of criminal punishment as a steady and unflagging progressive march of reform (e.g., Rothman 1971; Foucault 1977; Ignatieff 1978). These so-called revisionist accounts have been criticized largely for their overemphasis on the role of the state in shaping the nature of criminal punishments (Ignatieff 1981; Philips 1983). The civilizing process thesis places developments in the enterprise of criminal punishment in the context of other societal developments, which are less grounded in the logic of concrete agency but rather in the logic of discourse formation. Another equally important transformation in penal thought and practice can be detected in the transition to modernity. This shift concerns the focus of the penal sanction from the body social to the offender himself. In the deterrence paradigm that characterizes most systems of criminal justice prior to this time, the impact of penal measures on the offender is scarcely considered at all in formulating responses to his conduct; indeed, consideration of the offender’s motivation is largely absent. Implicit in this paradigm is the assumption that the potential for criminal conduct exists in every member of society. The objective of punishing the transgressor is therefore to prevent others who might be similarly tempted from following his example. The actual offender receiving the punishment is effectively written off in this process; society has no more business with him. Under the deterrence paradigm, the process of punishing the offender does not address him (or his conduct) as a member of society; punishment, in effect, can be seen as a process of turning away from the deviant. The “discovery” of the offender as an object of penal concern was crucial to the rise of the rehabilitative paradigm and of the penitentiary system that characterizes most modern penal systems.17 The increased use of transportation, penal slavery, and confinement thus represents the beginnings of a paradigm shift in the purpose and method of
Criminal Punishment in Civil Society
25
punishment. As McGowen (1995) has pointed out, “a [transportation] sentence of hard labor . . . had a double meaning, promising both suffering and reform” (McGowen 1995: 84). The idea of the offender as a principal target of penal intervention is perhaps the singular defining feature of the penal paradigms of the modern era.
THE EMERGENCE OF THE PENITENTIARY Incarceration as a major form of criminal punishment developed in England in response to the withdrawal of the American colonies as a destination for convicts as a result of the American Revolution. Although Australia became the principal destination for transported convicts after 1776 (Weidenhofer 1973; Shaw 1966; McGowen 1995), workhouses and “houses of correction” sprang up all over England, due in large part to the increasing problem of vagrancy (Peters 1995; Langbein 1998). Although deterrence still figured prominently in the goals of criminal punishment, the advent of the house of correction or bridewell in England marks the admission of other legitimate purposes to the punishment enterprise. These institutions were intended to “reform as well as punish” (McGowen 1995:83). Kuhn (1996) argues that paradigm shifts occur in response to crisis. The seeming ineffectiveness of deterrence in the late modern period (Hibbert 1963) and the loss of the American colonies as a destination for transported convicts, coupled with increasing resistance from Australian residents (Weidenhofer 1973) resulted in a renewed interest in the purposes of criminal punishment, what it could be expected to accomplish, and of the best methods for achieving its goals. A British treatise on penal measures written in 1863 indicates a new current of thought concerning the causes of crime and the relationship of punishment to reducing crime: “The sooner we learn that the crime engendered in England must also be dealt with in England the better. For as long as we can banish the monster evil to Australia, so long shall we neglect the means at our command for restraining it within the narrowest limits at home” (National Reformatory Union 1863:14). Similar rejections of the old paradigm and methods were pronounced in the United States, where imprisonment had come to replace many forms of punishment, but was still guided by the purposes of retribution and deterrence: Our penal codes assign imprisonment as a penalty for nearly every act they forbid, but . . . they leave it to the trial judge to fix the duration of imprisonment . . . according to his view of the criminal’s deserts. This system has often been exposed as absurd in principle and as grossly wrong and injurious in practice. It is founded on the false notion that the State can and ought to apportion retribution for offenses. . . . There
26
The Criminal in Society
are but two conceivable ways of protecting the community against its enemy, the criminal; to disarm him or to reconcile him. But the time sentence does neither. It restrains him until the term ends, as if one should cage a man-eating tiger for a month or a year, and then turn him loose. There is nothing in such a sentence which tends to reconcile him to his fellows. It commonly aims at nothing more than to restrain him and hold him safely for the term, and in most cases he is discharged more the foe of mankind than before. (Lewis 1863:59–60; see also Cary 1889:3–4) The notion that crime could be controlled by focusing on the treatment of offenders was central to the ascendance of the penitentiary. Some have credited nineteenth-century English penal reformer John Howard with refocusing the public discourse about crime and criminal punishment; Howard identified the causes of crime in society, rather than in individual failings, and had confidence in the prison as a disciplinary institution as a means of reforming criminal offenders (Hibbert 1963; McGowen 1995). Alternately, David J. Rothman (1995) argues that “a repulsion from the gallows, rather than any faith in the powers of the penitentiary itself, spurred the construction” (Rothman 1995:115) of prisons across Europe and America between 1780 and 1800.18 It is also the case that increasing population density (a function of both population growth and increasing urbanization) in both Europe and America posed logistic problems for the continuing use of banishment as a criminal sanction (McGowen 1995; Spierenburg 1995). In any event, it is clear that the emergence of the prison as the focal point of the enterprise of criminal sanctioning indicates an inchoate shift in the paradigm guiding the enterprise of criminal punishment. In the nineteenth century, incarceration became the primary method of punishment for criminal offenses in most Western nations. Some have argued for the importance of political economy in this development: the rise of the prison . . . was in significant part a product of the industrial revolution. It is not literally true that imprisonment for punishment . . . was unheard of in earlier times. Yet a system of long-term incarceration is an economic indulgence, and one beyond the means of most Western societies until near the end of the eighteenth century. (Allen 1981:12; see also Rusche and Kirschheimer 1939) While economic considerations undoubtedly had a great deal to do with the shift from transportation, galley slavery, and corporal punishment to imprisonment, changes in the prevailing ideologies concerning crime and the duties of a just society also played a significant role in this transformation. David Rothman (1971) argues that the rise of the penitentiary in nineteenth-century America
Criminal Punishment in Civil Society
27
can be attributed to changes in popular beliefs about the causes of crime, which motivated penal reformers of the time.19 Crime was believed to result from social dislocations accompanying rapid social change in postcolonial society, rather than from wickedness or sinfulness on the part of the individual: “Since the convict was not inherently depraved, but the victim of an upbringing that had failed to provide protection against the vices at loose in society, a well-ordered institution could successfully re-educate and rehabilitate him” (Rothman 1971:82; see also Wines 1871:560). The rehabilitative paradigm embodied in the institution of the penitentiary in America and Europe was compatible with the burgeoning “positivist” school of thought in criminology (Rotman 1990). The positivist school emerged in opposition to the classical criminology of the eighteenth century, which held that crime was the sole result of free choice on the part of the individual. Although the biological determinism embraced by Cesare Lombroso and his followers is commonly considered as the archetypal expression of the positivist school, many credit the pioneering work of Belgian statistician Adolphe Quetelet (1796–1874) as the intellectual founder of this school of thought ( Jones 1986; Beirne 1993). In examining criminal statistics in France, Quetelet observed an astonishing regularity in the amount of crime from year to year; he also found that certain social characteristics (e.g., education, religion, and age) covaried in a nonrandom fashion with criminal offending. Piers Beirne (1993) has characterized the impact of these discoveries on criminological thought thus: Quetelet’s insertion of criminal behavior into a formal structure of causality was a remarkable advance over the ad hoc and eclectic speculations of his contemporaries . . . within this formal structure, the shift of his analysis . . . allows him to claim that because crime is a constant, inevitable feature of social organization, it was “society,” “France,” or the “nation” itself that caused crime. (Beirne 1993:88)20
The Age of Rehabilitation The Jacksonian penitentiary differed from previous methods of prison confinement in that it was guided by a utopian vision of rehabilitation. The first American penitentiary was an annex to Philadelphia’s Walnut Street Jail, constructed in 1776. The penitentiary was a product of the efforts of the Quakers, who sought to curtail the use of corporal and capital punishments (Rotman 1990). In the Pennsylvania system, convicts served out their sentences in complete isolation, living and working with only the most minimal contact with other human beings. Once in isolation, it was argued, the prisoner “will be compelled to reflect on the error of his ways, to listen to the reproaches of his
28
The Criminal in Society
conscience”, and in this way be reformed (Rothman 1971:85; McKelvey 1936; Rotman 1990). Pennsylvania’s “separate system” was challenged by the advent of the “silent system” implemented at the penitentiary erected at Auburn, New York, in 1812. There, each offender slept in a private cell and worked by day beside other inmates, under a code of enforced silence (Rothman 1971). Despite minor administrative differences,21 both versions of the penitentiary system were founded on the same two premises: that the causes of crime were to be found within society, rather than in the individual, and that man was inherently perfectible and capable of being reformed. It must be noted, however, that the objective of deterrence was probably not insignificant in the shift to the penitentiary system in the United States: “undoubtedly some supporters were drawn to the program only because they believed that the terrors of isolation and silence would reduce crime” (Rothman 1971:89; see also McGowen 1995:97).22 The opening of the Auburn prison was followed by the construction of similar institutions in Massachusetts, Connecticut, New Hampshire, Vermont, Maryland, Kentucky, Ohio, and Tennessee (McKelvey 1936).23 It should be noted that while work was a conspicuous feature of the penitentiary system, this work is distinguishable from other forms of forced labor (e.g., slavery) in that it was not an end unto itself. Work was intended to aid in the rehabilitation of the offender, by teaching the values of discipline and endurance. Also distinguishing the penitentiary from other modes of punishment was the fact that confinement itself was instrumental to the primary objective of punishment. The reformulated notion of the criminal found expression in the “New Penology.” The most widely circulated treatise of the time was the proceedings of the National Congress on Penitentiary and Reformatory Discipline, held in Cincinnati in 1870. Penal reformers from all over the nation attended this meeting, at which a list of principles were adopted to define the new paradigm. These principles not only specified what the penitentiary was to be, but also what it was not: Punishment is suffering, moral or physical, inflicted on the criminal, for the wrong done by him, and especially with a view to prevent his relapse by reformation. Crime is thus a sort of moral disease, of which punishment is the remedy. . . . The treatment of criminals by society is for the protection of society. Since, however, punishment is directed, not to the crime but to the criminal, it is clear that it will not be able to guarantee the public security and re-establish the social harmony disturbed by the infraction, except by re-establishing moral harmony in the soul of the criminal himself. . . . The supreme aim of prison dis-
Criminal Punishment in Civil Society
29
cipline is the reformation of criminals, not the infliction of vindictive suffering. (Wines 1871:548) The indeterminate sentence was essential to the New Penology and to the rehabilitative paradigm it represented. Although there were minor structural differences in the application of indeterminate sentencing across different jurisdictions, there were several features that were common to all systems. The central idea of the indeterminate sentencing system was that offenders should be detained until such time as they would no longer be a threat to society; the burden of determining the offender’s fitness to rejoin society fell to a panel of experts administering a system of parole (Wines 1871; Zalman 1977–1978). Indeterminate sentencing as applied in the United States was not strictly indeterminate. Most jurisdictions specified a maximum term of imprisonment; at the discretion of the parole board; offenders could be released at any time before this if they were deemed to have been reformed. However, under the terms of this conditional release offenders could be returned to prison in the event of misbehavior for the duration of their term, or until parole authorities saw fit to release them at an earlier date (Zalman 1977). The rehabilitative paradigm makes the offender central to the process of punishment. Indeed, the individualized character of the indeterminate sentence places explicit focus on the treatment of the offenders while minimizing the salience of their transgressions (Wicharaya 1995). Offenders were to be released back into society at such time as they could be identified as rehabilitated. The penal innovation of indeterminate sentencing swept rather quickly through the nation; in 1877 Michigan passed the first indeterminate sentencing statute. By 1915, twenty-six states had such statutes in place (Miller et al. 1982), and by 1922, only four states had not adopted some form of indeterminate sentencing (Dershowitz 1976).
PENAL PURPOSE IN THE TWENTIETH CENTURY The Decline of Rehabilitation The “rehabilitative ideal” that was born in Philadelphia in the late eighteenth century held nearly undisputed sway until the 1970s (Allen 1981; Rothman 1971, 1980; Flanagan 1996). Increasing rates of crime throughout the 1960s and 1970s, as well as a “radical loss of confidence in [American] political and social institutions” (Allen 1981:18) resulted in a reshaping of the discourse surrounding imprisonment, particularly concerning the goals of incarceration. In addition to more “applied” investigations into the proper purposes of imprisonment, there
30
The Criminal in Society
was also something of a resurgence in scholarly treatments of the history and social meaning of the prison (e.g., Rothman 1971; Foucault 1977; Ignatieff 1978). The attack on rehabilitation came from many different camps. Some observers—both liberals and conservatives—expressed dismay at the arbitrariness and injustice they perceived as resulting from indeterminate sentencing systems (Frankel 1973; Morris 1974; van den Haag 1975; von Hirsch 1976). Others denounced the philosophical underpinnings of rehabilitation as essentially coercive, and as such antithetical to American ideals of liberty (Rothman 1971; Kittrie 1971; Mitford 1973; Morris 1974). The essence of a paradigm shift is summed up by Kuhn (1969) as follows: Successive paradigms tell us different things about the population of [a] universe and about that population’s behavior. . . . But paradigms differ in more than substance, for they are directed not only to nature but also back upon the science that produced them. They are the source of the method, problem-field, and standards of solution accepted by any mature scientific community at any given time. As a result, the reception of a new paradigm often necessitates a redefinition of the corresponding science. (Kuhn 1969:103) Kuhn also notes that the emergence of a new paradigm necessitates the rejection of the old one. It must not simply be shunted aside; it must be characterized as wrong (Kuhn 1996:115). In disciplines undergoing scientific revolutions, Kuhn also notes that paradigm shifts are rarely portrayed as such, but rather as linear progress toward the existing base of accumulated knowledge. While the rehabilitative ideal was denounced in many different arenas, perhaps the most prominent was the declaration of a panel of researchers who, having undertaken an exhaustive survey of rehabilitative programs and strategies, concluded that “nothing works”24 (Martinson 1974; Lipton, Martinson, and Wilks 1975). Another conspicuous rejection of the rehabilitative paradigm can be found in the text of the Comprehensive Crime Control Act of 1983, when the United States Congress officially disparaged the “outmoded nineteenth-century rehabilitative theory that has proved to be so faulty that it is no longer followed by the criminal justice system” (Congressional Information Service 1986). It is not simply that rehabilitation does not “work,” as the readers of Martinson and his colleagues might conclude; rather, the entire theory is flawed.25
In Search of a New Paradigm The complexities of the real world rarely mirror our theoretical models exactly. In our neatly constructed theoretical universe, the last forty years in American
Criminal Punishment in Civil Society
31
criminal justice would consist of a tidy succession of paradigm shifts, with incapacitation emerging as the undisputed winner. While I do contend that incapacitation has emerged as the dominant paradigm of criminal punishment at the end of the twentieth century, the years between rehabilitation’s demise and incapacitation’s reign appear as something of a hodgepodge of penal strategies in search of a rationale. However, despite the lack of an easily identifiable national reform movement like the one that was apparent in establishing the rehabilitative paradigm in postcolonial America, paradigmatic themes can be detected in this period.
Determinate Sentencing and “Neoretributionism” The first clear indication of the emergence of a paradigm shift away from the objectives of rehabilitation can be seen in the determinate sentencing movement that swept the nation in the 1970s. What was striking about the movement for determinate sentencing was the overwhelming lack of dispute about the benefits of substituting determinacy for the lengthy indeterminate sentences that were an essential part of the rehabilitative paradigm.26 Political liberals favored determinate sentencing as a response to sentencing disparity among offenders convicted of similar crimes (particularly as manifested along racial lines), and supported the reduction of discretion in a system that was viewed as capricious and often unnecessarily cruel (Kadish 1978; von Hirsch and Hanrahan 1981; Hewitt and Clear 1983; Rothman 1983).27 Determinate sentencing was also seen as more humane, in that offenders would be able to form realistic expectations about the probable length of their term of imprisonment, something that was not always possible under vague indeterminate sentences, which might read “one year to life imprisonment” (Rothman 1983). Conservatives favored determinate sentencing for its more manifestly punitive aspects. Conservatives also supported the elimination of sentencing disparities—but for different reasons. It was believed that the reduction of disparity in punishment would increase the certainty of punishment, thus enhancing its deterrent effect (e.g., Wilson 1975). This combination of utilitarian and desert-oriented aims have led John D. Hewitt and Todd R. Clear (1983) to label the paradigm that underlies determinate sentencing neoretributionism. The rhetoric of determinate sentencing certainly reflects a new kind of retributive philosophy—it consists of punitive claims cloaked in social defense justifications.While criminal punishment under the rehabilitative paradigm operated under a rubric of “treatment,” the determinate sentence has a different object toward which it is oriented. While the indeterminate sentence focused on the offender as a person, the determinate sentence addresses itself to the act. Offenders, in this scheme, are merely
32
The Criminal in Society
bearers of criminal acts, one interchangeable for another.28 This perspective is explicit in the stated goals of reducing sentencing disparity: The legislature finds and declares that the purpose of imprisonment for crime is punishment. This purpose is best served by terms proportionate to the seriousness of the offense with provision for uniformity in the sentences of offenders committing the same offense under similar circumstances. (California Sentencing Act, 1976, reproduced in Allen 1981:8) The overriding concern of determinate sentencing schemes is the quanta of punishment (von Hirsch and Hanrahan 1981), and not, as in the rehabilitative paradigm, the “experience of punishment that matters most” (Rothman 1983:634). Indeed, the nature of the experience of punishment and of institutions of punishment was rendered insignificant by the neo-retributionist paradigm. Determinate sentencing schemes turned sentencing into a “numbers game”; in this framework, all that was needed was to find the correct calculus of punishment to service retributive concerns and also to reap some utilitarian benefit through deterrence and incapacitation (Cassou and Taugher 1978).
The Renascence of Retribution: Upping the Ante Reflecting the new wave of neo-retributionist sentiment, deterrence enjoyed a brief period of theoretical prominence in American criminology during the period spanning the middle 1960s to the late 1970s, as seen in the veritable explosion of research on the topic during this period. The beliefs about the nature of crime and criminals embodied in the deterrence paradigm is also reflected in the work of prominent academic criminologists during this period, such as Lawrence E. Cohen and Marcus Felson (1979).29 In an explication of their routine activities perspective, they declare “we do not examine why individuals and groups are inclined criminally, but rather we take criminal inclination as given and examine the manner in which the spatio-temporal organization of social activities helps people to translate their criminal inclinations into action” (Cohen and Felson 1979:589). The profusion of empirical studies on the deterrent effects of incarceration ultimately provided only weak or equivocal support for the existence of such effects (see Nagin 1978 for a review of these studies). The overall review of the deterrence literature led Daniel Nagin to conclude in 1978 that Although more punitive sanctioning practices might legitimately be argued as a responsible ethical response to a truly significant crime
Criminal Punishment in Civil Society
33
problem, arguing such a policy on the basis of the empirical evidence [for deterrent effects] is not yet justified because it offers a misleading impression of scientific validity. (Nagin 1978:136)30 Sentencing policies in the 1980s reflected an increasing emphasis on retribution as the primary justification for criminal punishment. Several factors may account for this. The failure of empirical research to provide consistent support for deterrent effects undoubtedly influenced the increasing emphasis on the retributive elements embodied in the neo-retributionist paradigm; however, other, more subtle influences can also be identified. Although determinate sentencing schemes have taken a variety of forms (see von Hirsch and Hanrahan 1981 and Wicharaya 1995 for a review of the specific functional forms), with few exceptions, the return to determinate sentencing has resulted in increases in the rate of prison commitments.31 It has been argued that this is because under indeterminate sentencing schemes, judges were less likely to sentence those convicted of less serious offenses to prison when to do so might result in lengthy terms of incarceration (Blumstein 1983). With the shorter fixed terms prescribed in the new determinate sentencing structures, judges could be assured that this would not happen, and thus became more amenable to sentencing convicted offenders to prison. One consequence of this increased reliance on imprisonment is a process of escalation of the severity of criminal sanctions. When incarceration is the presumptively applied sanction for most crimes, the only way for legislators to satisfy constituencies’ demands to “get tough on crime” is to increase sentence severity. As Franklin E. Zimring (1984 [1977]) has observed: [R]eallocating power to the legislature means gambling on our ability to make major changes in the way elected officials think, talk, and act about crime. Once a determinate sentencing bill is before a legislative body, it takes no more than an eraser to make a one-year “presumptive sentence” into a six-year sentence for the same offense. (Zimring 1984 [1977]:273) Mandatory minimum sentences can be seen as the logical consequence of this bar-raising effect of determinate sentencing structures. The number of mandatory minimum sentencing statutes on the books increased throughout the 1980s (Caulkins et al. 1997). Mandatory minimum sentences are expressive of a more purely punitive paradigm than simple determinate sentences. Although mandatory minimums correspond to the social defense justifications of deterrence and incapacitation, the fact that so many of the mandatory minimum sentences instated in this period were for drug offenses speaks to their true paradigmatic significance.
34
The Criminal in Society
Retribution, the simplest penal paradigm, is also the most “emotional” or “nonrational” justification for imprisonment (Hewitt and Clear 1983). Diana R. Gordon (1994) has highlighted the moral nature of the “War on Drugs” campaign that commenced in the early 1980s. The moral character of the “war” has effected a paradigm shift toward a more purely retributive position. The symbolic nature of the drug war is demonstrated by the very endurance of antidrug policies in the face of little evidence that these strategies are successful in reducing either illicit drug use or crime (Trebach 1987; Gordon 1994; Baum 1996; Bertram et al. 1996; Caulkins et al. 1997). The forces underlying the paradigm shift toward a greater emphasis on retribution can be discerned in the work of many American criminologists in this period.32 During the 1980s and 1990s, several prominent works in the academic literature in criminology presented a quite different view of the criminal and causes of his behavior than those seen during the reign of the rehabilitative paradigm. In many ways, these authors represented a return to the neoclassical project of Gabriel Tarde (1912), whose work has been described by Piers Beirne (1993) as “an attempt to establish a compromised terrain between the unbridled subject of classical jurisprudence and the overdetermined object of positivism” (Beirne 1993:171).33 Exemplars of this view include James Q. Wilson and Richard J. Herrnstein’s Crime and Human Nature (1985) and Michael R. Gottfredson and Travis Hirschi’s A General Theory of Crime (1990). These works present the criminal as a failure of both morality and biology, offering a heady mix of culpability and inevitability that is remarkably compatible with the retributive paradigm. If, on the one hand, the criminal becomes so due to the choices he actively makes (or, in the words of Gottfredson and Hirschi, his lack of selfcontrol), this makes him worthy of punishment. However, if we can identify an organic condition (e.g., left-handedness, one of the correlates of criminal propensity identified by Wilson and Herrnstein) that may have predisposed him—although not compelled him—to choose crime, then there is little point in attempting to rehabilitate him. This late twentieth-century combination of classical conceptions of free will with just a touch of Lombrosian positivism thus abandons all humanitarian pretense; punishment of the offender expresses only our rightful outrage at his very existence.
Incapacitation: The Paradigm of Last Resort Zimring and Hawkins (1995) assert that incapacitation has emerged as the dominant paradigm in American thinking about prisons by default, due largely to the empirical failures of the rehabilitation and deterrence paradigms. They note that “the amount of scholarly attention devoted to the incapacitation process has been minimal . . . incapacitation can thus be regarded as the punishment objective of
Criminal Punishment in Civil Society
35
last resort” (Zimring and Hawkins 1995:158; see also Spelman 1994:vi). Actually, this is something of an overstatement. A considerable amount of attention in American criminology has been given to the notion of incapacitation as a penal purpose (e.g., Greenberg 1975; Cohen 1978, 1983; Greenwood with Abrahamse 1982; Gottfredson and Hirschi 1986; Nagin 1998). There are two general functional forms by which to achieve criminal incapacitation; collective and selective incapacitation (Greenberg 1975; Cohen 1983). Collective incapacitation entails increasing sentence severity for all offenders convicted of a designated offense, perhaps taking into account prior criminal history. Criminal sentencing under a policy of selective incapacitation is less influenced by conviction offense, and more dependent on some system whereby “dangerous” or “high-rate” offenders can be prospectively identified and detained for longer periods of time than other offenders who are similarly situated with respect to conviction offense, but who are not so identified.34 Although a great deal of scholarly research was devoted to estimating the likely impacts of incarceration strategies based on both collective and selective incapacitation in the 1970s and 1980s, the results were uniformly disappointing. In a comprehensive review of studies estimating the collective incapacitation effects of various sentencing schemes, Jacqueline Cohen (1983) found that the declines in crime rates that might be expected to result from toughening sentences varied (according to offense type targeted and study methodology) between 4% and 42%. However, she also found that existing sentencing structures were estimated to produce crime rate reductions ranging from 1% to 25%, and that proposed schemes to increase the effects of collective incapacitation would increase prison populations between 310% and 523% (Cohen 1983:12–31). The empirical prognosis for selective incapacitation was every bit as dismal as that of the prospects of collective incapacitation as a remedy for the problem of crime. In addition to the numerous ethical problems involved in sentencing similarly situated offenders differently based on predictions about their future behavior (see Auerhahn 1999a for a discussion of these issues), the evidence concerning the ability of criminal justice practitioners, researchers, and judges to prospectively identify “dangerous” or “career” offenders indicates that high rates of error characterize all predictive efforts to date (e.g., Steadman and Cocozza 1974; Thornberry and Jacoby 1979; Monahan 1981; Greenwood with Abrahamse 1982; Auerhahn 1999a). Simply put, the prospective identification of high-rate offenders appears to be beyond the current capabilities of clinicians, criminological researchers, and practitioners. Despite the lack of empirical confidence that can be mustered for policies based on the principle of selective incapacitation, the allure of this idea remains strong today. In the face of a well-established literature on “career criminals,” which offers fairly sound evidence that a small number of offenders contribute disproportionately to the volume of crime (Wolfgang, Figlio, and Sellin 1972;
36
The Criminal in Society
Petersilia 1980; Chaiken and Chaiken 1982; Wright and Rossi 1986; Shannon 1991), Americans appear to be stubbornly attached to the fanciful idea that the selective incapacitation of particularly dangerous offenders can provide a solution to all of our crime problems. In this sense, Zimring and Hawkins are correct in their assessment of the reasons for the prominence of incapacitation in penological discourse today; in explaining the ascendance of incapacitation as a penal paradigm in the 1990s, the apparent failure of the paradigms that followed upon the heels of rehabilitation’s demise emerges as a likely exegesis.35 The renewed interest in selective incapacitation in the 1990s has undoubtedly helped foster the rise of what has been called the “actuarial” school of criminology.36 This perspective is concerned with analyzing the growing influence of actuarial practices in crime control and in the management of criminal offenders (Reichman 1986; Simon 1987, 1993; Feeley and Simon 1992; O’Malley 1992; Rigakos 1999), but could also be argued to encompass the community of researchers and practitioners concerned with developing and implementing systems of actuarial classification for inmates (e.g., Hoffman 1992; see also Gottfredson and Tonry 1987). In an influential article, Malcolm M. Feeley and Jonathan Simon (1992) explore the consequences of actuarialism on incarceration policy and practice. They term this constellation of effects “the new penology”:37 The new penology is neither about punishing nor about rehabilitating individuals. It is about identifying and managing unruly groups. It is concerned with the rationality not of individual behavior or even community organizations, but of managerial processes. Its goal is not to eliminate crime but to make it tolerable through systematic coordination. (Feeley and Simon 1992:455) The popular currency of the selective incapacitation paradigm is apparent to even the casual observer of criminal justice policy today. The recent proliferation of “Three Strikes and You’re Out” habitual-offender statutes targeting “dangerous” offenders, now in existence in twenty-four states and the Federal System, can be seen as the logical culmination of the evolution of philosophies of incarceration into today’s focus on incapacitation combined with liberal doses of retribution.38 This ideological shift toward an emphasis on retribution and incapacitation is also reflected in recent public opinion surveys. In reviewing a number of studies, Flanagan (1996) cites evidence that “when provided with the social defense goal (incapacitation), Americans choose this objective over punishment by a wide margin” (Flanagan 1996:82). Similarly, Flanagan and S. R. Caulfield (1984) concluded from a review of public opinion surveys that the main function the public assigns to prisons is that of “protection of the public” (Flanagan and Caulfield 1984:4; see also Innes 1993). These researchers also found that,
Criminal Punishment in Civil Society
37
no matter how the questions are asked, public support for rehabilitation has consistently declined, while support for “protection of society” increased 166% in the period between 1968 and 1982 (Flanagan and Caulfield 1984). When a national sample of respondents was asked in 1995 whether governmental efforts should be targeted toward “rehabilitation” or “punishment” of criminals, 59% of respondents chose the “punishment” option, while only 27% identified “rehabilitation” as a worthy target of efforts (Gerber and Engelhardt-Greer 1996:72).39 Speaking to the pervasiveness of the ideological shift from rehabilitation to some combination of the objectives of incapacitation and retribution, Gerber and Engelhardt-Greer (1996) found that demographic factors were not significantly predictive of public attitudes about the goals of sentencing and incarceration.
CONCLUSION It has been said that “the criminal justice system has been burdened with unrealistic expectations of solving social problems that have proved insoluble elsewhere” (Griset 1996:127). The increased use of imprisonment in the latter half of the twentieth century can undoubtedly be attributed, at least in part, to this phenomenon. What I have attempted to do in this chapter is to explain why it is that penal reforms take particular forms by analyzing the expectations and beliefs of those advocating and implementing the reforms. I have argued that these expectations and beliefs crystallize into identifiable paradigms that convey the raison d’être of imprisonment in a particular time and place; these paradigms, which also encompass beliefs about the nature of crime and criminals, influence the types of sanctions that are chosen. Later chapters will be devoted to the examination of the consequences of those choices, both in terms of their intended objectives as well as unintended consequences. Toward that end, chapter 3 details the specific sentencing policies that have resulted from the expression of these paradigm shifts in the state that shall serve as the site of our empirical inquiry, California.
This page intentionally left blank.
Chapter 3 Criminal Sentencing Policy and Paradigm Change in California
The preceding chapter charted the historical trajectory of paradigm changes in American criminal justice. It was argued that paradigm shifts occur in response to operational or ontological crises in the criminal justice system—in other words, new paradigms rise to dominance when existing paradigms fail to provide satisfactory solutions to designated problems, or when the explanations of criminal behavior and the prescribed responses contained within a paradigm lose coherence. As chapter 2 demonstrated, these two circumstances often complement and reinforce each other. Chapter 2 described the timing and consequences of these paradigm shifts with respect to national trends in changes in sentencing structures. This chapter offers an account of the development of criminal justice paradigms in California, again with a focus on the changes in sentencing structures resulting from these paradigm shifs, and the consequences of these changes to the California criminal justice system. California is an interesting case in which to examine the historical movements in sentencing paradigms. Several scholars have noted that California often exemplifies national trends in criminal sentencing policy (Kadish 1978; Messinger and Johnson 1978; Tonry and Morris 1978; Casper, Brereton, and Neal 1983). For example, it has been observed that in California “the rehabilitative ideal produced the most conspicuously indeterminate and individualized prison sentences in the United States” (Tonry and Morris 1978:254). Similarly, the incapacitative paradigm that dominates American criminal justice in the 1990s was expressed in extremist fashion when, in 1994, the California electorate voted in the most far-reaching and widely implemented “Three Strikes” habitual-offender law in the nation. California is also useful for studying the relationship between paradigm change and sentencing reform due to the sheer 39
40
The Criminal in Society
volume of sentencing reform activity that has taken place in the latter half of the twentieth century. Since the passage of the 1976 Determinate Sentencing Law, the California legislature has enacted over 1,000 new laws relating to criminal sentencing policy (Foote 1993:8); in the 1990s alone, the legislature passed more than 400 bills increasing sentences (Schrag 1998). California also has an noteworthy and unique relationship with the paradigm whose consequences we shall later explore, selective incapacitation. While California sentencing policy is largely expressive of national trends, such as rehabilitation, neo-retributionism, and the like, the specific ways in which these paradigms have been interpreted in California are indicative of a consistent underlying concern with the incapacitation of dangerous persons. For example, when Folsom Prison was erected as California’s second state prison, it was designated by correctional authorities as the prison to which “hardened criminals” would be sent, so as to separate them and keep them from “contaminating” the other prisoners at San Quentin, who where ostensibly more amenable to reform. Similarly, the provisions of the Indeterminate Sentencing Law assured that truly dangerous offenders need never be released.1 Even California’s Determinate Sentencing Law, which was explicitly intended to remove disparities of punishment among offenders convicted of the same crime, allowed for a somewhat complicated system of “sentence enhancements” that allowed the judge to take into account the dangerousness of the offender by way of his or her prior convictions. Additionally, California retained the indeterminate sentence for a number of the most serious crimes, mostly those offenses involving interpersonal violence. Indeed, when examining California sentencing policy in historical perspective, one wonders if the state’s Three Strikes law, the consummate expression of the selective incapacitation paradigm in the 1990s, was not an inevitable consequence of what had come before.
SOURCES OF PARADIGM CHANGE IN CALIFORNIA CRIMINAL JUSTICE The conditions that give rise to paradigm shifts in criminal justice have multiple and diverse origins. While the process of paradigm change in California is not substantially different from that in the rest of the nation, an analysis of the main sources of change adds to our understanding of how paradigms interact with sentencing policy to produce stability or change in system goals and outcomes in a specific time and place. One major source of paradigm change in the history of the California penal system is popular opinion and the views that citizens hold about crime and criminals.2 California provides its citizens with a voice in governance via the ballot ini-
Criminal Sentencing Policy and Paradigm Change in California
41
tiative.3 Voted into law in California in 1911, the ballot initiative process exhibits a direct connection between public opinion and criminal justice policy in the state. In recent years, the number of ballot initiatives regarding crime and criminal justice has increased dramatically ( Jones 1998; see figure 3.1). In the face of the perception of increasing rates of crime, citizens are able to translate their fear of crime into legislation by way of the direct ballot initiative; two prominent examples of this are the 1982 “Victims’ Bill of Rights” and the Three Strikes statute that was overwhelmingly supported by California voters in 1994.4 In recent decades, in addition to the increase in the number of criminal justice-related initiatives, initiatives of all kinds have increased; additionally, the rate at which initiatives are approved has doubled, from approximately 22% of initiatives placed on the ballot garnering majority support to a high of 44% in the 1980s (Allswang 1991:13; Schrag 1998). The initiative process has become very important in recent years; Peter Schrag explains this by noting that the combination of the diversity in the state, the structure of the California legislature, which requires a two-thirds “supermajority” to pass budgets and most appropriation bills, and the strongly partisan nature of California politics creates a political environment in which elected officials do what [their] constituents want, but there may not be much that is both constitutionally permissible and that two-thirds of the people [legislators] can agree on. In the state’s disjointed system, it is sometimes structurally easier to get a majority of the voters than it is to get two-thirds of their representatives. (Schrag 1998:204)5 The process by which an initiative comes to be placed on the ballot is much more complicated than it appears at first glance. In order for a proposed initiative to qualify for placement on the ballot, a certain number of verifiable signatures must be collected in support of the initiative’s placement on the ballot.6 In the last two decades, the business of initiative qualification has indeed become a big business. One observer claims that “in some cases . . . it wasn’t clear whether it was the sponsor or the commercial consultants who had conceived and initiated the measure” (Shrag 1998:211). Despite the increasing commodification of the democratic process in the state, I contend that just as modes of punishment will only be employed if they are not fundamentally inconsistent with prevailing community sensibilities (see chapter 2:23–24), initiatives will only succeed in getting placed on the ballot and ultimately approved by a majority of voters if they are within the range of what is, to use David Garland’s (1990) phrase, “emotionally tolerable” (Garland 1990:214). An example of an initiative that failed to qualify for the ballot in 1994, the same year that voters overwhelmingly affirmed their support for the state’s Three Strikes law, makes this point. The measure would have provided for
42
The Criminal in Society
implementation, before release from custody, of facial identifying numbers on prisoners guilty of [a] felony in which a violent act is committed or threatened, or any crime while illegally in possession of any instrumentality which would provide means to commit great bodily injury, or property damage from which personal injury could reasonably, proximally result. Repeat offenders receive additional facial implants. . . . Establishes telephone number to provide information on wearer of implant number. (Secretary of State 1994)7 Paradigm change in California is also attributable to operational crises in the penal system itself. For example, as will be seen below, parole was initially introduced in response to facility crowding, and was not, as in many other states, an expression of the rehabilitative consensus. Parole did come to assume a meaningful place in the arsenal of rehabilitative technologies, but this was after it was already in place, having been established for pragmatic reasons. Another example of a crisis within the system that was related to paradigm change came about as the result of a widespread perception of disparity arising from the unfettered discretion of sentencing judges under the indeterminate sentencing system. The
FIGURE 3.1 Criminal Justice Ballot Initiatives Proposed in California, 1934–1998*
Sources: Friedrich (1999); Jones (1998). *The first criminal justice-related intiative was proposed in 1934.
Criminal Sentencing Policy and Paradigm Change in California
43
paradigm change that ushered in the determinate sentencing system can be seen as a rejection of the individualized model of rehabilitation in favor of a uniform scheme of retribution and deterrence. Many have noted the increasing scope of influence of the California Correctional Peace Officers Association (CCPOA), the prison guards’ union, as a force in criminal justice policy change (Schiraldi 1994; Davis 1995; Lotke 1996; Pens 1998). The CCPOA was one of the largest financial supporters of the Three Strikes ballot initiative campaign, and they regularly contribute to victims’ rights groups fighting to toughen criminal penalties ( Jones 1998; Schiraldi 1994). The gains that accrue to correctional officers from the expansion of the penal system are unmistakable. In addition to increased employment opportunities, there has been substantial wage growth within this occupation. The average salary earned by a California prison guard in 1980 was just over $14,000; today, the average salary is over $40,000 (Schiraldi 1995). The prison industry itself has been cited as a source of policy and paradigm change (Christie 1994). This is particularly evident in California, where twenty-one new prisons have been erected since 1984 (California Department of Corrections 1995, 2000); economically depressed rural communities often welcome prisons and the jobs they bring (Davis 1994; California Department of Corrections 1995; Lotke 1996; Gilmore 1998). In addition to the financial benefits realized by organized interest groups like prison guards’ unions, private corrections firms like Wackenhut and the Corrections Corporation of America as well as companies that provide services like telephone service and cable television to inmates also share in the profits that accrue from sentencing changes resulting in greater numbers being locked up for longer periods of time (Lotke 1996). The “corporatization” of the prison in California, as elsewhere, may encourage and support the growing emphasis on management strategies and efficiency in the enterprise of criminal punishment.
CRIMINAL JUSTICE IN CALIFORNIA: BEGINNINGS In 1850, California became the thirty-first state in the Union. This fact alone has some interesting consequences for the evolution of criminal justice ideology and policy, given that California entered into statehood when the rehabilitative paradigm was already firmly entrenched in much of the nation. It is important to note, however, that the cultural forces driving the rehabilitative consensus were largely centered in the eastern part of the country. For this reason, the timing of penal paradigm shifts in the West lagged behind those in the East in the nineteenth and early twentieth centuries (Bookspan 1991). This cultural gap closed with advances in transportation and communication in the later part of the twentieth century.
44
The Criminal in Society
The first prison in California, San Quentin, began receiving prisoners in 1854.8 From the beginning, the objective of this prison was primarily incapacitation. San Quentin in its early years has been characterized by one writer as “little more than a depository for the hated and unwanted” (Bookspan 1991:xviii). When the San Quentin location was proposed in response to an 1852 legislative resolution to erect a state prison, the subcommittee found the site “in every respect suitable for such a purpose . . . . the place is somewhat secluded and will be easily guarded” (reproduced in Bookspan 1991:6). Much of California’s early history is a story of virulent nativism and of the criminalization of the foreign-born (Starr 1873; McWilliams 1946; Boswell 1986; Guerin-Gonzales, 1994; Auerhahn 1999b). In light of this, it is not surprising that the majority of convicts housed in the State Prison at San Quentin were noncitizens; in 1858, 55% of San Quentin’s residents were foreign-born. In 1873, aliens were no longer a majority, but they still comprised 47% of California prisoners (Bookspan 1991:xviii).9 The state prison was also intended to curtail the activities of bands of vigilantes that had been active in the territory since the 1830s by establishing a climate of “law and order” (Bean and Rawls 1983). San Quentin’s early years were troubled. The management of the prison was originally under contract to several enterprising businessmen, who gained the right to use the labor of the convicts in exchange for securing and caring for the prisoners. After several years of mismanagement and widely publicized scandals involving alcohol, corruption, and brutality, the state regained control of the prison in 1858 (Lamott 1961; Owen 1988; Bookspan 1991).
THE BEGINNINGS OF PENAL REFORM: TOWARD THE TWENTIETH CENTURY The “New Penology” of eastern reformers was likely introduced to California by way of the 1870 “Declaration of Principles” drafted at the National Congress on Penitentiary and Reformatory Discipline (Bookspan 1991). The Reverend James Woodworth, a prominent California penal reformer, was in attendance (Wines 1871). After the 1870 meeting, Woodworth embarked on a quest to establish an Elmira-style reformatory in California at San Quentin. His dream was finally realized, albeit in limited fashion, with the opening of the branch prison at Folsom in 1880. The new prison was designated as the facility for “hardened” criminals, while San Quentin was to house younger offenders serving shorter sentences (Bookspan 1991). Parole release was introduced in California in 1893. Parole was not, however, a construction of rehabilitation-oriented reformers; it was rather a mechanism devised by legislators to remedy two operational problems endemic to the California penal system. One was the massive sentencing disparities result-
Criminal Sentencing Policy and Paradigm Change in California
45
ing from unfettered judicial discretion in sentencing. In different courts in the state, similarly situated offenders would receive vastly different sentences—for the same crime, one might receive a sentence of two years, another ten (Messinger et al. 1985). The other problem was prison crowding. In arguing to the legislature in 1893 for the institution of a parole system, Governor Henry Markham noted that California prisons at the time housed from two to three times as many prisoners as in any other State in the Union in proportion to population. . . . I believe it is due to two reasons. First, our statutes create such an exceedingly large number of State prison offenses. Second, because the Judges of this State, in their discretion, impose excessive sentences as compared with other states. (Markham 1893: 44–45, cited in Messinger et al. 1985:83) Parole, then, provided California with a mechanism for granting clemency on a wide scale, something that far exceeded the capacity of the existing apparatus for sentencing review. It was not something that was initially considered as a tool for use in the rehabilitative enterprise. It was an extremely pragmatic solution to California’s perennial problem, overcrowding.10 However, the institution of parole relates to the rehabilitative ideal in a special way in California. Since parole was necessary to relieve prison crowding, correctional authorities had a stake in making parole politically palatable to the general public. The success of a parole system depends on the reintegration of the parolees into the community. In furthering this objective, correctional authorities promulgated the idea of the criminal “having paid his debt to society” and of the prison as a corrective institution (Whyte 1916; Messinger et al. 1985:99). Bookspan (1991) contends that the ethos of individualism was much more influential in the rise of the rehabilitative paradigm in California than were the communitarian ideals that dominated eastern penal discourse.11 Eastern penal reformers often spoke of their “duty” to the criminal; since society had failed him, he was deserving of the rehabilitative efforts of the prison. In California, penal reformers believed that every man was capable of reform, not necessarily that society owed him the chance to reform. The western take on rehabilitation was largely unconcerned with the “external” (i.e., those outside the individual) causes of criminal behavior. This individualist stance resulted in a vision of rehabilitation that was oriented more toward social defense concerns than the humanitarian motives that were behind the “New Penology” in the East (Bookspan 1991). The primary beneficiary of rehabilitation was not the offender, who after his release would go on to live a more fulfilling life, but rather the society that was to receive him, reform having conquered his predatory impulses. The 1910 report of the State Board of Prison Directors illustrates both the individualist and social defense elements of the California rehabilitative paradigm:
46
The Criminal in Society
It would be impossible, within the confines of this report, to set forth all the methods and details of management that may be advantageously employed in a reformatory for adult offenders, nor can such methods be rigidly prescribed. Each prisoner presents a separate and independent problem, and every problem has a human and personal element. . . . [I]t must be kept in mind that each prisoner is a man, subject to certain defects which it is the purpose of the institution to correct. . . . [T]he final purpose of the prison is not to humiliate or degrade, but rather to elevate and reform. The average criminal has a distorted view of his relations toward organized society. It is to reverse and correct this view that reformatory and corrective institutions are established. (State Board of Prison Directors 1910:6–7; emphasis added) The prominence of social defense objectives within the rehabilitative paradigm in California is apparent in the classification system that developed with the opening of Folsom Prison in 1880. From the beginning, Folsom was intended to house the “hardened criminals,” while San Quentin would become a reformatory prison for younger and seemingly more redeemable offenders. In the penal discourse of California, a branch prison was necessary for rehabilitation to be successful, given the explicit recognition that there was presumed to exist— either by choice or by destiny—a class of permanently dangerous offenders.
THE INDETERMINATE SENTENCING LAW AND THE REIGN OF REHABILITATION IN CALIFORNIA California’s Indeterminate Sentencing Law was enacted in 1917. As in other state systems, the indeterminate sentence was intended to serve as a tool in the rehabilitation of offenders. The ever-present promise of release encouraged reform, while indeterminacy assured that the unrepentant could be detained for long periods of time. Under the law, judges had broad discretion to set prison terms. Offenders would often be sentenced within extremely wide ranges, such as a sentence of six months to life. Messinger and Johnson (1978) point out that sentencing judges did not set prison terms at all; judges merely decreed that the convicted offender be sentenced to the “term prescribed by law” (Messinger and Johnson 1978:15). The offender’s actual release date was to be determined by the parole board (in 1944 the Adult Authority was established to perform this function), a group of experts who “fixed” the convicted offender’s term at some point between the statutory minimum and maximum. The term established by the Authority represented the total length of time that the offender was to be under the supervision of the state. A parole date, some time short of the maximum term, was also set at this time. Under the Indeter-
Criminal Sentencing Policy and Paradigm Change in California
47
minate Sentencing Law, prisoners served, on average, about one-third of the minimum term; offenders given life sentences were usually eligible for parole in seven years (Messinger and Johnson 1978). The Department of Corrections was established in 1944 in response to concern with mismanagement and conditions of confinement in California prison facilities. Governor Earl Warren organized an investigatory committee and concluded from their findings that “the conditions . . . found to exist in our penal system are a challenge to every public spirited citizen of this state. The solution lies in a complete reorganization of this function of state government” (California Department of Corrections 1995:8). No historical account of the California penal system would be complete without some discussion of Richard A. McGee. In 1944, after a nationwide search, McGee was appointed the first director of the newly created Department of Corrections. Prior to his appointment, McGee had been an elementary schoolteacher, and held a number of administrative posts in the Federal Penitentiary System (Glaser 1995). McGee’s vision shaped the penitentiary system and the approach to the transformative enterprise of rehabilitation in California. He held a strong commitment to the rehabilitative ideal that he carried with him throughout his career.12 It is widely agreed that conditions in California prisons improved substantially under McGee’s direction (Glaser 1995:21–27). He was thoroughly committed to the rehabilitation of offenders, and pioneered a number of innovations designed to accomplish this goal. These innovations included changes in personnel selection, training, and promotions; in addition, McGee’s Department of Corrections established a centralized base of authority, and broke down the “fiefdoms” of the wardens of individual institutions (Glaser 1995: 31–36). Under McGee’s direction, the Department also implemented numerous inmate programs, including vocational training, alcohol and drug abuse programs, individual and group counseling, and a variety of educational programs (Glaser 1995:43–82). The era of the rehabilitative paradigm as expressed in the Indeterminate Sentencing Law was one of relative political tranquillity. The rehabilitative consensus appears to have been firmly entrenched throughout the population. One indication of this is the utter absence of grassroots reform activity. For a period of nearly forty years (1936–1974), not a single ballot initiative relating to crime or criminal justice was proposed (see figure 3.1). This stands in stark contrast to the profusion of activity in this arena in the postrehabilitative era; in the 1980s alone, more ballot initiatives pertaining to crime were proposed than had been throughout the seventy-nine years prior to 1980 that the initiative process had been in existence in California ( Jones 1998). Although rarely used prior to the implementation of determinate sentencing in California, the ballot initiative was used on one occasion to
48
The Criminal in Society
reform the criminal justice system in a way that demonstrates the reach of the rehabilitative paradigm. Placed on the ballot of November 5, 1974, Proposition 10 sought to amend the state constitution to restore the right to vote to ex-felons upon the completion of their prison or parole terms.13 It was approved by a majority of voters, who apparently agreed with the measure’s proponents that “an ex-felon returned to society and released from parole has fully paid the price society has demanded . . . the objective of reintegrating ex-felons into society is dramatically impeded by continued restriction of the right to vote.”
THE DEMISE OF REHABILITATION: PARADIGM CHANGE AND THE DETERMINATE SENTENCING LAW Throughout the 1960s and 1970s, Americans became increasingly dissatisfied with the criminal justice system and seemed to lose confidence in its ability to shield society from the dangers of crime. Softhearted judges were blamed for meting out “revolving-door justice” (Kadish 1978). David J. Rothman (1983) has identified one source of this dissatisfaction with discretion in criminal justice within a larger societal trend—namely, the distrust of discretion as a means for producing equitable outcomes, and the rejection of “experts” of all kinds. There appear to be shades of this phenomenon in the debates about sentencing policy in California during this period. For example, one critique of California’s Indeterminate Sentencing Law refers derisively to the “almost godlike power” of sentencing judges (McGee 1974a:4). The passage of the Determinate Sentencing Law (DSL) in 1976 was preceded by a growing consensus about the deficiencies of California’s indeterminate sentencing structure. Part of this discontent was manifested in a new vision of the criminal. The political and academic discourse throughout the 1960s and 1970s resulted in the gradual replacement of the view that “every man has within him a germ of goodness” (cited in Bookspan 1991) with the image of a sullen miscreant who failed to be rehabilitated, at least in part due to his “unwillingness . . . to take advantage of such opportunities as the state does provide” (McGee 1974:6; see also Howard and Hugh 1978). The growing interest in victims’ rights also fueled the movement toward greater determinism in sentencing and away from rehabilitative concerns. Under the tenure of Ronald Reagan, this interest was codified into the establishment of a “victims’ awareness week” (Ellingwood 1985). California’s Uniform Determinate Sentencing Act of 1976 became law on July 1, 1977. The act was largely the result of two pieces of legislation: Senate Bill 42 and Assembly Bill 476 (commonly referred to as the Boatwright Amendment).14 The move toward greater determinism in sentencing was widely
Criminal Sentencing Policy and Paradigm Change in California
49
supported by those interested in reducing disparity, as well as by those aiming to increase the deterrent and retributive efficacy of sentencing (Carey 1979). The text of the act explicitly articulated the parameters of the new paradigm: The Legislature finds and declares that the purpose of imprisonment for crime is punishment. This purpose is best served by terms proportionate to the seriousness of the offense with provision for uniformity in the sentences of offenders committing the same offense under similar circumstances. (Section 1170 reproduced in California District Attorneys’ Association 1979: VI-1) The Determinate Sentencing Law left a good deal of discretion in the hands of the sentencing judge. It did not specify whether or not offenders convicted of particular crimes should be sentenced to prison; it merely specified term lengths for particular offenses if the judge chose a prison term from the array of available sanctions. Indeterminate sentencing was retained for some crimes (e.g., murder) and still persists today. It is not uncommon for offenders convicted of murder or nonnegligent manslaughter to receive wide-ranging sentences such as fifteen years to life (Leonard 2002). For most offenses, the DSL provided for a limited range of terms, specifying a low, middle, and high value (e.g., 3, 4, or 5 years). The middle term was presumptively applied; departures from this required written justification on the part of the sentencing judge explaining the reasons for the mitigation or aggravation of sentencing severity (Nagin 1977). In addition to the base terms prescribed by the legislature, judges could add various sentence enhancements, for instance, for use of a firearm in the conviction offense (1 to 3 years, depending on whether or not the firearm was used, and the consequences to the victim). Other enhancements could be given for multiple conviction offenses, excessive property damage occurring as a result of the offense, and prior convictions (Messinger and Johnson 1978; Nagin 1977).15 The DSL also allowed for up to one-third of the sentence to be reduced by the accumulation of “good time” credits. These good time credits served two purposes. For one, while good time was “vested” and after a certain point could not be taken away from an inmate, it could fail to be awarded and as such provided a check, albeit weak, on inmate misbehavior. More importantly, good time credits preserved some of the “safety-valve” function previously performed by discretionary parole. The primary consequence of the implementation of the DSL in California was an increase in prison populations (Brewer et al. 1981; Austin and Pannell 1985; see figure 3.2). This was due to both increased rates of prison commitments and an increase in the average length of stay. The increase in commitments came about as a result of judges’ increased willingness to sentence “marginal” defendants to a term of incarceration, now that they were assured
50
The Criminal in Society FIGURE 3.2 California Prison Population, 1925–1996
Sources: Langan et al. (1988), Bureau of Justice Statistics (1999).
that there was no way that such a defendant would serve an unduly long sentence (Corrections Digest 1980; Brewer et al. 1982; but see also Casper, Brereton, and Neal et al. 1982). The causes of the increased length of stay were a bit more complex. The statutory base terms established by the DSL were based on average terms served under the indeterminate systems. However, the provision for a variety of sentence enhancements added extra time on top of these “average terms”—into which, under the old system, such aggravating circumstances were already taken into account at the time of the Adult Authority’s fixing of the term. To make matters worse, the effect of longer average terms could not even be offset by parole. Under the DSL, discretionary parole release gave way to “supervised release” for a period of not more than three years following release from prison (Carey 1979). As early as 1977, Franklin E. Zimring high-
Criminal Sentencing Policy and Paradigm Change in California
51
lighted the importance of parole in regulating a criminal justice system driven by legislative sentencing reform. Parole is an important “safety valve” that serves to maintain correctional populations at a manageable level while still retaining the symbolic benefits of “harsh” sentencing policies.
SOWING THE SEEDS OF VENGEANCE: MANDATORY MINIMUMS AND RETRIBUTION IN CALIFORNIA Mandatory sentences have been a feature of California sentencing policy for many years. The legislature placed limits on the range of sanctions available to sentencing judges as early as 1975, with a statute that prohibited a probation sentence when a defendant was convicted of certain offenses using a firearm (Ashman 1979). In 1979, mandatory minimum sentences for certain drug offenses (S.B. 469) and violent offenses (S.B. 406) were instituted by the California legislature (Kannensohn 1980). In keeping with the national trend, mandatory minimum sentences increased in California throughout the late 1970s and 1980s. The passage of the DSL represented an important step in the direction of increased punitiveness in California sentencing policy. Indeed, this trend had already been established before the DSL; a study by Berk, Brackman, and Lesser (1977) of changes in the California Penal Code in the postwar period found that reforms were almost invariably in the direction of greater punitiveness. Through the use of the ballot initiative, California voters instituted a number of “gettough” and mandatory sentencing reforms in the years immediately following the passage of the DSL. In 1978, Proposition 7 increased penalties for defendants convicted of first- and second-degree murder, and expanded the “special circumstances” required for death penalty eligibility.16 Proposition 4 on the June 1982 ballot placed restrictions on bail release for defendants, changing the status of bail from a presumptive right to a discretionary privilege. The most farreaching of these get-tough sentencing reform initiatives was 1982’s Proposition 8, dubbed by its proponents “The Victim’s Rights Initiative.” Proposition 8 amended the state constitution to provide for the mandatory use of sentence enhancements for prior felony convictions, and also placed limitations on both bail and parole release (Assembly Committee on Criminal Justice 1982). The amendment also established programs for offenders to provide restitution for victims of crime. California’s long-standing interest in the incapacitation of dangerous persons is reflected in the drive for increased punitiveness. While all incarceration serves an incapacitative function to a greater or lesser extent, collective incapacitation is expressly recognized as a benefit of increasingly punitive sentences. The following is taken from the text of the amendment created by Proposition 8 to Article 1, Section 28 of the state constitution:
52
The Criminal in Society
The rights of victims pervade the criminal justice system, encompassing not only the right to restitution from the wrongdoers for financial losses suffered as a result of criminal acts, but also the more basic expectation that persons who commit felonious acts causing injury to innocent victims will be appropriately detained in custody, tried by the courts, and sufficiently punished so that the public safety is protected and encouraged as a goal of highest importance . . . [T]o accomplish these goals, broad reforms in the . . . disposition and sentencing of convicted persons are necessary and proper as deterrents to criminal behavior and to serious disruptions of people’s lives. (Assembly Committee on Criminal Justice 1982:2) The push to “get tough” in California has resulted in an expansion of the penal system the likes of which is unprecedented anywhere in the world. In an attempt to keep up with exponential prison population growth, California voters approved ballot initiatives that issued an average of $500 million in general obligation bonds for prison construction every two years between 1982 and 1990. Taking into account indirect costs such as interest, the bond initiatives approved for prison construction since 1984 will cost Californians over $5.2 billion (Davis 1995). Twenty-one new facilities have already been constructed since 1984, and at least eight more are in the planning or construction stages (Pressman and Kaae 1996); it is estimated that the state would need to construct fourteen new prisons just to relieve existing overcrowding (LAO 1997). One author has wryly observed, “it is hard to drive California’s freeways nowadays without coming to signs showing an exit to some correctional facility” (Schrag 1998:97). The paradigm shift highlighting the punitive aspects of penal institutions is indicated in the names of the new facilities (see table 3.1). Prior to the 1980s, facilities had euphemistic names such as the “California Men’s Colony,” “California Rehabilitation Center,” or “California Correctional Institution.” The facilities built as part of the construction program following the era of determinate and mandatory sentencing have less lofty aspirations, boasting more utilitarian appellations like “Wasco State Prison.” The debates surrounding the bond issues for new prison construction are in fact more illustrative of consensus than of conflict. Although the question of whether each specific initiative ought to be adopted or not is hotly contested by those on both sides, little disagreement emerges about the value of the laws creating the need for more space to house increasing numbers of prisoners. Rather, the debates focus around differences of opinion about strategies for financing new prison construction. No one argues against the construction of more prison cells, and no one speaking out (on either side) about these measures expresses dismay about prison population increases. In urging voters to ap-
TABLE 3.1 California State Prison Facilities
Facility
Date Opened
Facility
Date Opened
San Quentin Folsom Prison1 Women’s Prison at Tehachapi2 California Institution for Men Deuel Vocational Institution Correctional Training Facility3 California Medical Facility California Institution for Women California Men’s Colony California Correctional Institution California Rehabilitation Center4
1852 1880 1932 1941 1946 1946 1950 1952 1954 1955 1962
California Conservation Center5 California State Prison, Solano6 California State Prison, Sacramento7 Avenal State Prison Mule Creek State Prison8 Richard J. Donovan Correctional Facility Northern California Women’s Facility9 California State Prison, Corcoran Chuckawalla Valley State Prison Pelican Bay State Prison Central California Women’s Prison10
1963 1984 1986 1987 1987 1987 1987 1988 1988 1989 1990
Facility
Date Opened
Wasco State Prison California State Prison, Calipatria California State Prison, Los Angeles County North Kern State Prison California State Prison, Centinela Ironwood State Prison Pleasant Valley State Prison High Desert State Prison Valley State Prison for Women Salinas Valley State Prison
1991 1992 1993 1993 1993 1994 1994 1995 1995 1996
Sources: California Department of Corrections (1995, 2000). Notes: 1
Folsom was authorized by the California legislature in 1858; construction did not begin until 1874, and the prison received its first inmates in 1880 (CDC 1995). This facility was severely damaged in an earthquake in 1952; the inmates were housed in tents on the grounds until the replacement prison, the California Institution for Women, opened later that year (CDC 1995). 3 This is the prison at Soledad, made (in)famous by George Jackson’s writings (1970). 4 This facility was originally designated to serve the state’s civil addicts program. 5 The name of this facility was changed in 1973 to the California Correctional Center. 6 The Solano facility was originally administered as part of the California medical facility; the two prisons split in 1992 (CDC 1995). 7 The Sacramento State Prison was originally set up as an annex to Folsom; the prison acquired a separate warden in 1992 (CDC 1995). 8 A report from the California Department of Corrections (1995) notes that “citizens of the nearby town, Ione, lobbied to have this prison built” (26). 9 Martin Roth (1993) notes of this prison that “the first inmates were received July 27, 1987, and design capacity was reached on August 12, 1987” (Roth 1993:217). 10 Currently housing over 3,100 inmates, this is the largest women’s prison in the world (CDC 2000). 2
54
The Criminal in Society
prove the $495-million bond issue of the New Prison Construction Bond Act of 1981 (1982’s Proposition 1), the proponents boast that Since 1975, California has enacted more tough anticrime legislation than at any other time in the state’s history. Prison is now mandated for many major crimes . . . California has made it clear that convicted criminals will go to prison. As a result, many more criminals are going into the prison system each month. In fact, prison commitments have doubled in the last seven years, resulting in more than 30,000 inmates in California prisons, the largest number in our history. (California Ballot Pamphlet 1982a:6) The thrust of the concerns expressed by the opposition is expressed in the rebuttal to this argument: “There is no question about increasing prison populations or the need for additional facilities . . . California needs prisons, but Proposition 1 is too expensive and financially unsound” (California Ballot Pamphlet 1982a:6). In the June primary election of 1984, the debate over Proposition 17 merely reprised these positions. Once again, the arguments against the $300-million bond issue centered around fiscal responsibility, and did not articulate any coherent objection to the expansion of incarceration. Indeed, throughout the 1980s, the only opposition that was mounted against the campaigns to support the increased incarceration resulting from punitive sentencing reforms focused on “cutting the fat” in the corrections bureaucracy, and mismanagement of the state budget. Indeed, the argument offered in the ballot pamphlet for the election of November 4, 1986 assures voters that “[s]upporters of Prop. 54 say that without this bond no new prisons can be built to lock up criminals. That is POPPYCOCK! Prisons will be built if Californians defeat this bond” (California Ballot Pamphlet 1986:11, emphasis in original). Although California voters continued to approve bond issues for prison construction until 1992, the beginnings of the shift toward a greater emphasis on the selective incapacitation of dangerous offenders are evident in the arguments surrounding construction proposals as early as 1988, when opponents to Proposition 80 argued that there is something wrong when the population grows only 21% while the number of alleged felons grows 257% over a ten-year period . . . we do not believe that tens of thousands of evil felons have moved to California in the last ten years. (California Ballot Pamphlet 1988:14) In the 1990 primary election, arguments on both sides of the prison construction bond issue on the ballot reflected a strong emphasis on incapacitation. For
Criminal Sentencing Policy and Paradigm Change in California
55
example, supporters of the measure urged voters to pass the measure because “Proposition 120 will provide the funds needed to continue building more prisons so that we can remove dangerous criminals from your neighborhoods and keep them behind bars where they belong” (California Ballot Pamphlet 1990: 54). In the same vein, opponents argued that “the proponents of Proposition 120 would have you believe that these cells are needed to protect you from dangerous criminals. In fact, fewer than half of the inmates in prison today have been convicted of violent crimes” (California Ballot Pamphlet 1990:54).
SELECTIVE INCAPACITATION AND “THREE STRIKES AND YOU’RE OUT ”: CALIFORNIA LEADS THE WAY On March 7, 1994, California’s “Three Strikes and You’re Out” habitual-offender statute took effect. The “Three Strikes” Movement was promoted by victims’ rights advocates in the state, most notably Fresno resident Mike Reynolds, whose eighteen-year-old daughter Kimber was attacked and murdered in 1992 by two parolees. The gunman was subsequently killed in a shoot-out with police, and the other offender received a prison sentence of only nine years. Outraged, Mike Reynolds approached two democratic assemblymen (Bill Jones and Jim Costa) from Fresno, who drafted the first Three Strikes bill. The bill was defeated in committee, and Reynolds began a campaign to place the measure on the ballot as an initiative statute. Support for the measure heightened in October 1993, after twelve-year-old Polly Klaas was abducted from her bedroom in a middle-class suburb of San Francisco and was murdered by Richard Allen Davis, a man with a lengthy criminal history who had been released from prison. The Polly Klaas case became a focus of the public outcry to do something about “career criminals.”17 The state legislature hurriedly passed A.B. 971 on March 3, 1994, and Governor Pete Wilson signed the bill into law on March 7, 1994. Proposition 184, the ballot initiative originated by Mike Reynolds, remained on the November 1994 ballot despite the fact that it was nearly identical to the law signed by the governor eight months earlier. One reason given by supporters of the proposition was that passage of the initiative would prevent the legislature from weakening the statute with legislative amendments (Pollard 1994:4). Despite the lack of concrete information about the probable impacts of the statute, the measure was popular with the California electorate, and was affirmed by more than 70 percent of the voters. Under the provisions of California’s Three Strikes law, upon a second conviction for any of the “serious or violent felonies” (named in Sections 667.5 and 1170.12 of the Penal Code, and Section 707[b] of the Welfare and Institutions Code), the offender is subject to an automatic doubling of the presumptive sentence for the conviction offense, plus any relevant sentence enhancements.18
56
The Criminal in Society
The Third Strike provision of the law requires that offenders with two prior convictions for “serious or violent felonies” convicted of any felony be sentenced to three times the normal presumptive term (to be served consecutively), or twenty-five years to life, whichever is longer. Additionally, offenders sentenced under either the second or third -strike provisions must serve 80 percent of their sentences before becoming eligible for release due to “good time” credits. Selective incapacitation and retributive ideals were prominent in California elections throughout the 1990s. An example of both of these can be seen even earlier, in Proposition 89 (1988), in which the voters granted the governor the power to veto parole release decisions for convicted murderers. What is most striking about this particular initiative is the utter lack of organized opposition to the proposition. Unlike other criminal justice policy ballot initiatives, where arguments represent the views of coalitions of interests (and are signed by a number of representatives), the only argument offered in opposition to this reform was offered by a lone prisoners’ rights advocate. The measure was ultimately approved by a majority of California voters.
PROBLEMS AHEAD? California’s Three Strikes statute is intended to provide for the incapacitation of dangerous persons by targeting “habitual criminals” for mandatory sentences of twenty-five years to life. These habitual criminals are identified by the commission of three felony offenses. There are inherent problems with this as a strategy for selective incapacitation, not the least of which include the limitations of retrospective identification. Three Strikes defendants tend to be about ten years older than the average age of those committed to prison, and are therefore closer to the point of “aging out” of criminal activity, limiting the expected return on crime reduction to be derived from their incapacitation (Austin et al 1999). Another consequence of incarcerating middle-aged offenders for terms ranging upward of twenty years is the projected increase in the proportion of elderly offenders. Due to the relatively greater health care needs of these offenders, the annual cost of housing an elderly (over sixty years of age) offender is estimated to be at least twice that of housing a nonelderly inmate (Zimbardo 1994). Others have argued that the law will create problems for law enforcement officers as offenders become more determined to evade capture in the face of a likely thirdstrike charge; one such observer has noted that assaults on prison guards have nearly doubled since the law’s implementation in 1994 (Roemer 1996). The Three Strikes movement did not begin in California. Washington state approved a Three Strikes measure in 1993; in the past five years, twentythree states and the federal government adopted similar laws (Clark, Austin, and Henry 1997).19 However, as is characteristic of the state, California’s ex-
Criminal Sentencing Policy and Paradigm Change in California
57
pression of the national trend is exemplary. A report from the state’s Legislative Analyst’s Office declared the Three Strikes measure “the most significant change to the state criminal justice system in more than a generation” (LAO 1995). The Three Strikes statutes enacted in many states were largely symbolic, in that they tend to replicate already existing penalty structures and as such are rarely invoked (Butterfield 1996; Clark et al. 1997; Litvan 1998). California’s law, by contrast, has already been used to sentence nearly fifty thousand offenders (LAO 1999). There are two features of the law that largely explain this. The first is the “second-strike” provision, which mandates doubling the presumptive sentence for a second-felony conviction. These second-strike offenses account for the majority of Three Strikes prison admissions (Austin 1998). The second reason is that California’s “strike zone” is more inclusive than that of any other state, and includes many common felonies, such as drug violations and residential burglary. Additionally, the ballot initiative that affirmed the Three Strikes law was drafted such that it requires a two-thirds majority in both houses of the state legislature, or an initiative statute approved by the voters to reform the law. A related sentencing innovation that serves the objectives of selective incapacitation is Truth in Sentencing (TIS). TIS places limits on the accumulation of good-time credits and parole eligibility. As part of the 1994 Omnibus Crime Control Act, the federal government provided for “incentive grants” for states that were willing to bring their sentencing structures in line with TIS guidelines. Under TIS in California, offenders convicted of certain felonies20 must serve at least 85 percent of their sentences before becoming eligible for release. All of the problems in the California criminal justice system cannot be attributed to Three Strikes. While prisons represent the greatest proportional expense, there has been astronomical growth in all forms of criminal justice system supervision. Due to a combination of the structural fiscal limitations placed on state government coupled with the enormous expense of housing a prison population that increases, on average, 8 percent per year (Schrag 1998), the criminal justice system is ill-equipped to handle this growth. For example, The number of probation officers has remained about the same in the past twenty years, but the number of probationers has increased from an estimated 30,000 to 400,000, and each worker’s “intensive caseload” has gone up from fewer than 25 to close to 100. “You’re lucky,” said Susie Cohen, the executive director of the California Parole, Probation, and Correctional Assocation, “if you get a postcard once a month.” (Schrag 1998:99)21 Peter Schrag cites another example of the consequences of these fiscal constraints on criminal justice supervision at the local level: in Los Angeles
58
The Criminal in Society
County, where jail overcrowding is a perennial problem, increased by the advent of Three Strikes (Turner 1999), “a new $373 million jail stood empty for more than a year because the county couldn’t afford to run it” (Schrag 1998:99).
CONCLUSION This chapter has examined the effects of paradigm shifts in criminal justice policy in California. California is often regarded as a bellwether for the nation on policy issues; in the case of criminal justice policy, the state often expresses national trends in exemplary fashion. The cumulative result of these trends has been an unprecedented expansion of the state’s penal system. This expansion has grave consequences, particularly given the strains that supporting one of the largest penal systems in the world places on state resources, both fiscal and moral. This set of circumstances is generally regarded as an unintended consequence of the implementation of the Determinate Sentencing Law. The early evidence regarding the effects of California’s Three Strikes law indicates that the consequences of this particular reform may be far-reaching and at best, irrelevant to the objectives of the law. Certain of these consequences, such as the incarceration of large numbers of nonviolent offenders for lengthy terms, may be diametrically opposed to the primary objective of this and other recent reforms—namely, the incapacitation of dangerous offenders. The remainder of the book will be devoted to the empirical examination of the consequences of criminal sentencing reform with respect to the objective of incapacitating the dangerous. Chapter 4 examines the concept of selective incapacitation more closely, while chapter 5 examines the notion of dangerousness and dangerous offenders, central to any discussion of selective incapacitation. Chapter 6 presents a strategy for evaluating the success of reforms intended to selectively incapacitate dangerous offenders in California.
Part II Incapacitation and Dangerousness
This page intentionally left blank.
Chapter 4 Selective Incapacitation
In the previous chapter, I asserted that recent sentencing innovations in California were reflective of a growing interest in the idea of selective incapacitation. In this chapter, I examine the history of selective incapacitation as an idea, as well as the theoretical, operational, and ethical issues encompassed within it. Particular attention will be given to the 1982 Rand report, Selective Incapacitation, as this represents the most complete formulation of these ideas. The incapacitative rationale for incarceration is arguably the simplest from a theoretical standpoint: if an offender is isolated from society, then he or she cannot commit crimes in the larger society (Packer 1968).1 While simple theoretically, incapacitation is problematic to put into practice. While incarcerating everyone convicted of a crime for long periods of time would undoubtedly reduce the volume of crime committed by the offenders in question, research estimates indicate that a strategy of collective incapacitation would prove to be prohibitively expensive. For example, Jacqueline Cohen (1986) found that while increasing sentence lengths for all convicted robbers sentenced to prison might potentially reduce robbery rates by as much as 25%, the operational realities of implementing these sentences could increase prison populations by more than 500%. Selective incapacitation, then, represents an attempt to efficiently use the restraining capacities of incarcerative sanctions to ensure the public safety. Although the term selective incapacitation was first coined by David Greenberg in 1975, the most comprehensive proposal for a sentencing structure based on the principles of selective incapacitation is that offered by Rand Corporation researchers Peter W. Greenwood and Allan Abrahamse in their eponymous 1982 report. In this report, they proposed that a predictive scale 61
62
Incapacitation and Dangerousness
be used to distinguish between low- medium- and high-rate offenders at the time of sentencing, and that based upon these predictions, high-rate offenders be given longer sentences relative to their low-rate counterparts. Greenwood and Abrahamse contended that such a policy would reduce crime while using penal resources in an efficient manner to selectively incapacitate only high-rate offenders for long periods of time. While the particulars of the Greenwood and Abrahamse proposal have not been implemented in any jurisdiction, the influence of their ideas has been pervasive. A recent example of the popularity of selective incapacitation as a sentencing strategy can be seen in the proliferation of “Three Strikes and You’re Out” habitual offender statutes across the nation (Benekos and Merlo 1995; Clark, Austin, and Henry 1997). Preventive detention is an explicit goal of Three Strikes laws. Three Strikes statutes have as their basis the idea that “habitual criminals” identified by these laws will continue to commit crimes if not restrained from doing so (Schichor 1997).
“CRIMINAL CAREERS” AND “CAREER CRIMINALS”: THE BIRTH OF AN IDEA The premise upon which the selective incapacitation scheme is based is a body of literature on “career criminals,” a concept that originated largely from the work of Philadelphia cohort study researchers, who found that in a cohort of boys born and raised in the city of Philadelphia, 6.3% of the cohort (18% of all offenders) were responsible for 51.9% of all offenses committed by the cohort (Wolfgang, Figlio, and Sellin, 1972:89). This finding was replicated in another cohort study in Racine, Wisconsin (Shannon 1991). The 18% figure has also been closely approximated in studies of incarcerated populations. Chaiken and Chaiken’s (1984) analysis of the same self-report Rand data used in the Greenwood and Abrahamse analysis determined that 15% of the sample were highrate “violent predators.” Similarly, Wright and Rossi (1986) reported that out of a national sample of incarcerated felons, 22% of offenders (classified as “predators” by the researchers) were responsible for over 50% of the total criminality of the sample.2 Investigations into the nature of the criminal career are generally in agreement that the frequency of criminal offending declines with age (Petersilia 1980; Gottfredson and Gottfredson 1992; Nagin 1998; but see also Figlio 1996), and that there is little to no evidence of a progression of increasing severity of the offenses committed over the length of the “career” (Gottfredson and Gottfredson 1992; Shannon, 1991; Wolfgang, Figlio, and Sellin 1972; but see also Chaiken and Chaiken 19843) . Another finding common to studies of the career criminal phenomenon is that there is little to no evidence of specialization on the part of high-rate criminals (Wolfgang et al. 1972; Petersilia
Selective Incapacitation
63
1980; Wright and Rossi 1986; Gottfredson and Gottfredson 1992). As one group of researchers put it: it is also clear that the Predators [22% of total inmate sample] were not criminal ‘specialists’ in any sense of the word. They were, instead, what we can refer as ‘omnibus felons’—men prone to commit virtually any crime available in the environment for them to commit. (Wright and Rossi 1986:76)
SELECTIVE INCAPACITATION AND THE PROBLEM OF PREDICTION At first glance, the findings from criminal career research are logically consistent with the idea of selective incapacitation as an efficient means of targeting penal resources. However, what advocates of selective incapacitation usually fail to take into account is that researchers have also found that the proportion of offenders who cease criminal activity is constant at about 20–30% after each subsequent offense, and that this “dropping out” is a stochastic process (Wolfgang, Figlio, and Sellin, 1972: chapter 11). What this means is that while we can predict with some confidence that a certain proportion of offenders will be “career criminals” or “high-rate offenders,” prediction at the level of the individual offender is very difficult. Joan Petersilia (1980) concluded, based on her review of career criminal research, that “the data accumulated to date on criminal careers do not permit us, with acceptable confidence, to identify career criminals prospectively or to predict the crime reduction effects of alternative sentencing proposals” (Petersilia 1980:322). Morris and Miller (1987) identify three logical frameworks for predicting future behavior. Anamnestic prediction is based on past behavior; predictions of this type are predicated on the assumption that individuals are likely to behave the way they have in the past. Actuarial or statistical predictions are those that are based on previously determined estimates of risk for people who share some measurable attribute with the individual who is the subject of the predictive exercise. The use of demographic characteristics (e.g., age or race) in predictive models is an example of this type of prediction.4 Clinical prediction is based on individualized assesments of risk by experts of one sort or another. Psychiatric assessments of dangerousness are examples of clinical prediction. Up until the mid-1970s, clinical predictions were by far the most commonly used in the criminal justice system (Austin 1983). Errors in identifying high-rate offenders take two forms. Errors of underprediction are those cases that are predicted to be low-rate but, where crime is actually committed at higher rates, or false negatives; if these individuals are sentenced to relatively shorter terms based on this erroneous prediction, they
64
Incapacitation and Dangerousness
will be released sooner than they might have been under a desert-oriented sentencing policy, thus increasing the public’s risk of criminal victimization. Even more troublesome from an ethical standpoint are errors of overprediction, whereby offenders are erroneously predicted to be high-rate (false positives). False positives raise the obvious ethical problem inherent in incapacitating an individual to prevent him from committing crimes she would probably not have committed anyway. False positives are also problematic from a technical point of view, in that the resources devoted to incarcerating these incorrectly identified offenders for longer terms are not being put to their most efficient use. The scope of the impact of errors in predicting offense rates is not insignificant. Most schemes that purport to predict offender risk have false-positive rates of greater than 50% (Monahan 1981; Cohen 1983; Blackmore and Welsh 1983; von Hirsch and Gottfredson 1984). Most discussions of errors in prediction focus on the problem of false positives. This does not merely reflect a bias in considering one type of error more egregious than another; rather, this concern is well-founded, in that false-positive predictions are more common than false negatives in predicting a rare outcome (Copas 1983). The likelihood of a successful prediction depends on the base rate of the phenomenon in the population being studied. In the case of high-rate offending, the relatively small percentage of high-rate offenders (from 6% to 22%, depending on the population from which the sample is drawn) makes their identification via statistical prediction very difficult. This problem is exacerbated when the criterion behavior is extremely rare, as is the case with violence (Monahan 1981; Shah 1981; Steadman and Cocozza 1974). Conversely, it is very easy to correctly predict a behavior with a high base rate. To illustrate this point, two classification tables are presented (see tables 4.1 and 4.2). The first is from Greenwood and Abrahamse’s 1982 study. The second is from a replication of their analysis. Greenwood and Abrahamse performed their analysis on data collected in the late 1970s; the replication utilized a subset of the 1991 Survey of State Prison Inmates (see Auerhahn 1999a for details). The differences in the composition of the two samples with respect to the distribution of actual offending rates illustrates the consequences of base rates on predictive success. The base rates for high-, medium-, and low-rate offending are shown in the last row of both tables. The most striking difference between the two samples are the base rates for low-rate offending (53% of the 1991 sample vs. 30% of Greenwood and Abrahamse 1982), and high-rate offending (15.8% vs. 28%). The consequences of this are borne out in the analysis; the instrument was successful in identifying low-rate offenders 88% of the time when the base rate was 53%, but could only do so 52% of the time with a base rate of 30% in the Greenwood/Abrahamse sample. Similarly, the relatively low base rate of 15.8% for high-rate offending the replication lowered predictive accuracy in
Selective Incapacitation
65
TABLE 4.1 Predictive Accuracy of Greenwood/Abrahamse Scale Actual Offense Rate Predicted Offense Rate
Low
Medium
High
Total
Percent correct
Low Medium High Total
14% 12% 4% 30%
10% 22% 10% 42%
3% 10% 15% 28%
27% 44% 29% 100%
52% 50% 52% 51%
Source: Greenwood with Abrahamse (1982).
TABLE 4.2 Predictive Accuracy of Replication Scale Actual Offense Rate Predicted Offense Rate
Low
Medium
High
Total
Percent correct
Low Medium High Total
36% 12% 5% 53.0%
4% 14% 13% 30.7%
1% 5% 10% 15.8%
41% 31% 28% 100%
88% 45% 36% 60%
Source: Auerhahn (1999a).
identifying this group to 36%, as compared to the 52% accuracy exhibited in the Greenwood/Abrahamse sample, with a higher base rate of 28%.5 The consequence of the problem of limited predictive accuracy for selective incapacitation proposals was aptly summed up by Alfred Blumstein: Any stochastic sequence of events with a non-zero probability of termination after an event will inevitably result in a distribution of sequence lengths. In criminal-career terms . . . since every statistical distribution has to have a right-hand tail, the group of “chronic offenders” who comprise the right-hand tail will necessarily account for a disproportionately large number of offenses. The critical question is whether the members of this group are distinguishably different. . . . The fundamental policy question, then, is whether the “chronic offenders” are identifiable in prospect, that is, during the period in which they accumulate a record . . . unless such discrimination can be made,
66
Incapacitation and Dangerousness
any identification of chronic offenders can only be made retrospectively, and so is of little policy or operational value. (Blumstein, 1979; reproduced in Petersilia, 1980:374). The Greenwood/Abrahamse report represents the most detailed selective incapacitation proposal to date, and exhibits an overall success rate of only 51% in correctly classifying offenders using a seven-item predictive scale. The falsepositive rate (offenders erroneously classified as high-rate) for the analysis as reported by Greenwood and Abrahamse is 48% (1982:59);6 subsequent reanalyses of these data by Cohen (1983) and Visher (1986) arrive at an even higher false-positive rate of 55%. A recent analysis containing a replication of the Greenwood/Abrahamse model underscored the inadequacies of the predictive construct, exhibiting an overall accuracy rate of 60% and a false-positive rate of 64% (Auerhahn 1999a).7 The ethical problem posed by preventive detention in selective incapacitation schemes would be extant even if predictions were completely accurate, a fact noted by Chaiken and Chaiken (1982) in their analysis of the Rand self-report data utilized in Greenwood’s Selective Incapacitation: Even if the models were foolproof and the official records sufficiently complete and detailed, the legal and ethical ramifications of their use by the criminal justice system would be a matter of dispute. Sentencing offenders for past crimes which have never been adjudicated runs counter to principles of just deserts, while sentencing them for predicted future crimes runs counter to tenets of free will and justice. . . . Therefore, we suggest that our findings should not be used simplistically as criteria for passing judgment on specific individuals. (Chaiken and Chaiken, 1982:180) Preventive detention is defined as the confinement of persons based on a prediction of future dangerousness (Packer, 1968; Dershowitz, 1973; Morris and Miller, 1987). One of the most extreme historical examples of this practice is the internment of Japanese-Americans by government authorities during World War II; another example is the juvenile justice system’s practice of confinement for status offenses8 (Dershowitz 1973; Chesney-Lind and Shelden 1992). The selective incapacitation framework punishes/incapacitates offenders for crimes not yet committed, based on a prediction of future offending. Although there is little existing case law that addresses this issue directly, legal scholars are generally in agreement that preventive detention is inconsistent in principle with the foundations of American criminal law (see Packer 1968:93–103; Dershowitz 1973; von Hirsch 1984). Some have argued that predictions are used at every point in the criminal justice system—predictions that are made on an ad hoc basis by police, prose-
Selective Incapacitation
67
cutors, judges, and parole boards, and that statistical methods may improve the accuracy of these predictions (Gottfredson and Gottfredson 1985; Morris and Miller 1987). Zimring and Hawkins (1995) also note that all theories of punishment, save for a strict just deserts model, have an implicit predictive component. Greenwood and Abrahamse (1982) offered a similarly grounded defense to anticipated objections to their model: “the only alternative to preventive detention is a pure just deserts model, which rests on principles that are at odds with what the public seems to want and how the system currently operates” (Greenwood and Abrahamse 1982:92). Concerning the consequences of errors in prediction, Greenwood and Abrahamse (1982) respond thusly to the issue of the possibility of unjust dispositional outcomes on the basis of false-positive predictions: under a policy of selective incapacitation, some of the offenders who would be characterized as high-rate offenders and sentenced to longer terms would not actually have high offense rates. This possibility may offend some who would apply the same standards required for conviction—proof beyond a reasonable doubt—to the identification of highrate offenders. Nevertheless, for a number of reasons, the concept of selective incapacitation should not be immediately judged categorically unacceptable on ethical grounds. . . . It should be remembered that the model defined in this report should not be tested against completely accurate predictions, which we can never have, but against the current system. (Greenwood with Abrahamse 1982:27, 92) Even if one were willing to accept this argument, subsequent analysts of the Greenwood/Abrahamse selective incapacitation proposal have concluded that the predictive instrument does not improve on existing methods of identifying dangerous offenders (Cohen 1983; von Hirsch and Gottfredson 1984; Visher 1986), despite the authors’ claims (Greenwood with Abrahamse 1982:29). Others have attempted to justify selective incapacitation on the grounds that it is compatible with desert, in that offenders with lengthy criminal histories are comparatively more “blameworthy” for a given offense (Moore et al. 1984). However, this argument is problematic from a desert standpoint, in that high-rate offenders identified on the basis of self-reports or arrest data are at risk for receiving punishment for crimes that have never been adjudicated, which contradicts the basic tenets of due process embodied in American criminal law (von Hirsch 1985).9 Another set of objections that has been levied against the use of predictive instruments to identify high-rate offenders concerns the construction of these composite measures. Many items used in the construction of such measures, such as drug use and employment history, are unrelated to either the offense for which the offender is to be sanctioned, or the “blameworthiness” of the offender.
68
Incapacitation and Dangerousness
Indeed, the items comprising the predictive scale used by Greenwood and Abrahamse in their 1982 proposal were criticized for being little more than “proxies for race and class” (Blackmore and Welsh 1983; Cohen 1986). Additionally, composite scales are often poorly constructed from a methodological standpoint, thus severely compromising their reliability. For example, a recent analysis found that the seven items comprising the scale utilized by Greenwood and Abrahamse were so weakly interrelated that a measure of internal scale reliability (Cronbach’s alpha) did not meet the standard generally agreed upon for acceptance of a multiple-item scale in social science research (Auerhahn 1999a). Other problems with the use of predictive measures include the problem of shrinkage. Shrinkage is the term used to refer to the discrepancy between the goodness of fit (accuracy) of a predictive measure in the sample upon which the measure was developed and other samples (also called “validation samples”). A predictive measure will almost always exhibit a higher degree of predictive accuracy in the sample that was used to construct it. This is due to features of commonly used statistical estimation procedures, which tend to maximize the impact of any unique features in the sample data (Copas 1985). The effects of shrinkage are particularly distressing in the context of identifying dangerous offenders, given the extremely low levels of predictive accuracy that such instruments exhibit in their own construction samples. In an analysis of different types of predictive models in criminology, Gottfredson and Gottfredson (1992) found that measures designed to identify high-rate offenders exhibited high degrees of shrinkage, and were “among the least robust of those examined” (Gottfredson and Gottfredson 1992:iv).
ESTIMATES OF INCAPACITATION EFFECTS The allure of selective incapacitation derives from the reduction in crime ostensibly offered by such a strategy. Greenwood and Abrahamse estimated that the implementation of their proposal could reduce the robbery rate in California by 15%, while simultaneously reducing the population of incarcerated robbers by 5%. Estimates of the potential benefits of selective incapacitation vary widely (Blumstein, Cohen, and Nagin 1978; Spelman 1994; Zimring and Hawkins 1995). Crime rate reductions are most commonly considered as a summary function of the estimated offense frequencies of the individuals incapacitated, multiplied by length of sentence; this figure represents all the crimes that presumably would have been committed by these individuals, if free to do so (Blumstein, Cohen, and Nagin 1978; Greenwood with Abrahamse 1982; Cohen 1983).10 Estimates of individual offending frequencies () also exhibit a great deal of variability, with reported values ranging from 2 to 187 offenses per year (see
Selective Incapacitation
69
Spelman 1994:71–80 for a comprehensive review). Researchers estimating generally assume stable rates of offending over the duration of the career, and that all offenders have an equal probability of being arrested, convicted, and incarcerated (Greenwood with Abrahamse 1982; Cohen 1983; Spelman 1994). It is also assumed that sanctions have no impact on the criminal career; periods of incarceration are generally viewed as “interruptions” in the offenders’ career, which continues as before upon the end of a term of incarceration (Cohen 1983). The measure of central tendency used to represent the distribution of values is of crucial importance in determining the estimate of the aggregate incapacitative effect of imprisonment. The use of the mean results in a significant upward bias in estimates, due to the extreme skewness of the distribution of offense rates. The median is therefore a much more appropriate measure, although this is rarely the method used (Visher 1986). More often seen is the use of a censored or Winsorized mean (Greenwood with Abrahamse 1982; Spelman 1994). Whatever the method used, it is clear that the right-hand tail of the distribution of individual values should not be taken into account in formulating summary measures (Canela-Cacho, Blumstein, and Cohen 1997). Other factors that may account for some of the variability in estimates of offending rates are the types of samples used in generating estimates (Spelman 1994). All available evidence indicates, not surprisingly, that incarcerated samples are biased in the direction of having more high-rate offenders than are present in the general population (Wolfgang, Figlio, and Sellin 1972; Chaiken and Chaiken 1984; Wright and Rossi 1986; Shannon 1991; Canela-Cacho, Blumstein, and Cohen 1997). For this reason, estimates of derived from such samples are bound to exaggerate the potential impacts of selective incapacitation proposals. For example, Visher (1986) demonstrates that the values obtained by Rand researchers for the data used by Greenwood and Abrahamse suffer from a systematic upward bias due to the inflation resulting from cases with short street time, as well as from the method used to impute missing data. The level of sensitivity to initial assumptions in generating estimates is demonstrated in Visher (1986). Visher recomputed estimates of for subgroups (by state and offender type) in the Rand data. While the methods used in computations differed only slightly from those used by Greenwood and Abrahamse, the estimates obtained by Visher were uniformly lower than those presented in the original analysis—in some cases, these estimates were reduced by a factor of three. The work of Canela-Cacho, Blumstein, and Cohen (1997) demonstrates the consequences of the assumption of offender population homogeneity. Their findings suggest that the likelihood of being incarcerated varies directly with offending frequency, resulting in what Canela-Cacho, Blumstein, and Cohen call “stochastic selectivity.” They conclude that we may be at the upper limit of what we can accomplish in terms of crime reduction via incarceration:
70
Incapacitation and Dangerousness
In all analyses, the concentration of high- offenders found among inmates results entirely from stochastic selectivity operating on heterogeneous distributions of , and not from any policies to explicitly identify and incarcerate high- offenders . . . compared to inmates, free offenders [including those under other forms of criminal justice supervision such as probation or jail] are predominately low- offenders, Even though stochastic selectivity will continue to draw new inmates disproportionately from the high end of the distribution of free offenders, those new inmates will average lower s than current inmates, and their incarceration will reduce fewer crimes than the average for current inmates. Analyses of the impact of new incarceration policies that rely on mean s of prison inmates . . . substantially overstate the likely crime reduction to be derived from expanding imprisonment. (Canela-Cacho, Blumstein, and Cohen 1997:153–157) Estimates of the benefits to be derived from selective incapacitation policies are similarly sensitive to several other assumptions. These include not only the estimate of used in calculating crime rate impacts (and the assumptions embedded therein), but also the average career length imputed to offenders. Estimates of incapacitation effects based on an assumption of a five-year criminal career duration will differ greatly from those assuming a career length of twenty years (see Spelman 1994).11 Another problematic assumption in the calculation of incapacitation effects is that of a lack of replacement effects—that is, crimes attributed to incapacitated offenders are presumed to be prevented by their incarceration and subtracted from the total crime rate. However, Zimring and Hawkins (1995) note the importance of estimating both individual and community effects. Estimates of crime rate reductions should take into account not only the expected offense frequencies of incapacitated offenders, but also subsequent crime rates in the community from which the offender is removed. Zimring and Hawkins argue that community-based estimates of the crime rate impacts of selective incapacitation policies are likely to be lower than those based on individual offense frequencies, due to substitution effects and the effects of criminal groups (see also Blumstein et al., 1978:65; Spelman, 1994). In an incisive analysis, Zimring and Hawkins (1988) illustrate some of the problems inherent in the enterprise of estimating the potential crime-reducing effects of changes in sentencing policies. They used as their example Edwin W. Zedlewski’s 1987 National Institute of Justice report which, based on a value equal to 187 offenses yearly per offender, contended that the “social costs” averted through incapacitation exceeded the financial cost of incarcerating offenders by a factor of 17.12 While citing other deficiencies of Zedlewski’s analysis (such as his failure to consider the factor of diminishing marginal returns as imprisonment policies dip deeper and deeper into the offender pool), Zimring
Selective Incapacitation
71
and Hawkins also present some calculations using Zedlewski’s own estimates and assumptions. In so doing, they show that The total volume of crime estimated by the methods used in “Making Confinement Decisions” [Zedlewski 1987] was about forty million in 1977. At 187 crimes per criminal, the incarceration of about 230,000 extra offenders should reduce crime to zero on incapacitation effects alone. The problem is that on this account, crime disappeared some time ago because the U.S. prison population expanded by a total of 237,000 from 1977 through mid-1986. (Zimring and Hawkins 1988:428–429)
WHAT THE ESTIMATES DON’T TELL US Aside from the operational limitations of estimating incapacitation effects, enumerated above, there is a fundamental conceptual flaw in the strategy of computing crime rate reductions based on . Such estimates are based on the presumption that the high-rate offenders are successfully targeted by the sentencing mechanisms under evaluation. Let us consider the merits of this assumption. Even aside from the evidence presented above concerning the fallibility of existing methods to prospectively identify such offenders, is there any justification whatsoever for leaving this assumption unexamined? I will argue that there is not, and that in fact the examination of this assumption can provide us with an alternate means by which to evaluate the efficacy of sentencing policy reforms motivated by the idea of selective incapacitation. Consider this: it is entirely possible that a particular policy intended to target high-rate offenders might be quite successful in doing so, yet might not result in any reduction in the crime rate. Remember that the simple summation and subtraction method traditionally employed by researchers does not allow for the possibility that offenders are replaced by new offenders as they are removed from the community, nor does it consider the possibility of criminal groups such as gangs or organized drug distribution networks, the activities of which might well continue relatively undisturbed despite the removal of some members. It is possible that these effects could completely offset any benefit derived from the incapacitation of successfully targeted offenders. Examination of the assumption of selective success leads to new approaches in the empirical evaluation of selective incapacitation-driven policy innovations. It also raises an analytical question as well: How are we to define the “success” of a selective incapacitation policy? The traditional method of evaluating such policies exclusively in terms of crime rate impacts defines “success” in this limited fashion. But can sentencing policies really be expected to
72
Incapacitation and Dangerousness
effect such far-reaching change in the larger community? Pamala S. Griset (1996) has noted that “the criminal justice system has been burdened with unrealistic expectations of solving social problems that have proven insoluble elsewhere” (Griset 1996:127). A more realistic evaluation of a policy based on selective incapacitation might focus on the selective success of the policy, rather than on the volume of crime presumably prevented via incapacitation. In chapter 3, I asserted that California has historically exhibited a unique preoccupation with the incapacitation of dangerous offenders. Over the last three decades, a great deal of sentencing reform has been implemented in the state; to varying degrees, these reforms have explicitly intended to accomplish the goals of selective incapacitation. However, in the replication of the Greenwood/Abrahamse study mentioned earlier, utilizing the same distributional criteria as the original researchers revealed that a sample of 1991 California prison inmates contained 53% low-rate offenders compared to only 30% in the sample used in the 1982 Rand report (Auerhahn 1999a). This would seem to suggest that the sentencing policy reforms of the 1970s and 1980s, in addition to spurring unprecedented growth in prison populations, have also resulted in an overall weakening of selection mechanisms. Redefining the criterion by which we consider a sentencing policy to be successful in terms of selective incapacitation necessitates a new approach to the evaluation of sentencing policy reforms. Rather than attempt to prospectively identify the potential effects of possible sentencing policies, we might be interested in evaluating the success of existing policies in terms of their success in incarcerating dangerous offenders. This approach leads to several questions. First and foremost, how are we to define “dangerousness?” And how do we go about devising a strategy to assess the efficacy of sentencing policies with respect to this reformulated criteria for the success of selective incapacitation? The following chapters will be devoted to answering these questions.
Chapter 5 Dangerousness
The limitations of the traditional selective incapacitation model detailed in chapter 4 suggest that the retrospective evaluation of such policies may be more beneficial than predictive approaches. Due to the problems inherent in the estimation of crime reduction effects, I contend that the evaluation of selective incapacitation schemes should be based on the success of such policies in terms of their immediate objective, the incarceration of dangerous offenders. But what does it mean when we say that an offender is “dangerous”? The dangerous offender has long been part of the discourse of crime, criminals, and punishment. This chapter first presents an examination of the nature of the concept of danger, then a brief history of the concept of the dangerous offender. A number of attempts to identify and control such offenders are also discussed.
RISK, FEAR, AND DANGER Before we can address the question of who or what is dangerous, it is helpful to consider the concept of danger itself. The first thing that must be recognized is that danger is not an objective reality; rather, dangerousness is subjectively determined based on an evaluation of risk. It is possible to objectively determine risk; a risk is merely the estimated probability that some harm will occur. Danger is thus a risk that is deemed to be intolerable—either as a result of the magnitude of the risk, the nature of the harm in question, or some combination of both. A variety of elements may be influential in determining when a risk becomes intolerable. The nature of the anticipated harm may be even more important than the absolute numerical probability of its occurrence in determining whether or 73
74
Incapacitation and Dangerousness
not something poses a danger. William W. Lowrance (1976) identifies several factors that may play a role in this decision; these include consideration of whether the consequences of the harm are reversible or irreversible, whether the risk is avoidable, and whether or not the risk is borne voluntarily (Lowrance 1976:87–94; see also Walker 1978). Floud and Young (1981) offer the insight that “fear converts danger into risk” (Floud and Young 1981:6). They also note that when the judgment is being made about persons, intent is an important element in determining dangerousness: “the prospect of death or injury suffered at the hands of another person arouses greater alarm than death or injury suffered as the direct result of their dangerous or irresponsible behavior” (Floud and Young 1981:7; see also Morris and Miller 1985:11).1 The subjective derivation of judgments of dangerousness from risk implies another quality of dangerousness—relativity. A risk is the estimated probability of some event occurring, usually expressed with reference to some specified period of time (e.g., over the course of one’s lifetime or within the next month). By definition, risk is relative. While the probability of causing harm may be different for two individuals, it cannot be determined with certainty that either individual will, in fact, cause harm—only that we believe that one is more likely than the other to do so. By the same token, dangerousness is not a discrete characteristic. Individuals (or situations, or motor vehicles, or whatever is being judged dangerous) can properly only be deemed more or less dangerous than their counterparts.2 Unpredictability is a cardinal feature of dangerousness. This elusive quality is central to the fear that danger arouses in us: “If an ‘attack’ of dangerous violence can be anticipated and aborted, or treated, then it ceases to be dangerous” (Scott 1977:128; see also Lowrance 1976).3 Similarly, Pratt (1997) describes this quality of the dangerous offender as “unknowability”: “It was not only his repeated crimes . . . but this quality of unknowability as well that placed him beyond the norms of modern society and its available apparatus of penal control” (Pratt 1997:29–30; see also Foucault 1988:126–127). The essence of dangerousness is potentiality. All judgments of dangerousness are inherently forward-looking (Rennie 1978). When we speak of danger, our point of reference is not the past or present, but rather, we refer to the belief that harm will occur in the future. Similarly, when we speak of a dangerous individual, what we really mean is that we believe that at some point in the future, this individual will cause some harm that we have previously deemed to be intolerable.4
WHO DECIDES? The interactional quality of dangerousness is captured in Theodore Sarbin’s (1967) oft-quoted aphorism: “Violence denotes action; danger denotes a rela-
Dangerousness
75
tionship” (Sabin 1967:285). As the previous discussion indicates, danger is not something that has an objective existence; something only becomes dangerous when it is so designated. But who makes this designation? Whose definitions are privileged? In large part, the answer to these questions depends on the purpose of the definition. In the present context, dangerousness is considered as a means of evaluating the success of sentencing policies based on the logic of social defense via selective incapacitation. Therefore, the simplest criterion for establishing dangerousness would be the likelihood of an individual committing criminal acts of which the public is fearful in the future. Several authors have characterized the designation of dangerousness as a political issue (Rennie 1978; Monahan 1981; Pratt 1997). Lowrance (1976) describes the determination of safety (and therefore danger) as “a normative, political activity” as opposed to the outcome of a scientific process (Lowrance 1976:76). Rennie (1978) points out that “the answer to the question ‘who is dangerous?’ may well depend upon who is answering” (Rennie 1978:xvi). Those individuals and groups with access to power and resources are also likely to be able to define what is dangerous (Turk 1976, 1982); in the words of Ysabel Rennie (1978), “who has reason to feel threatened?” (Rennie 1978:55). An extensive literature exists in support of the notion that social control mechanisms are frequently mobilized in response to symbolic threats (e.g., Harring 1977; Irwin 1985; Jackson 1986; Myers 1987, 1989, 1990, 1993; Brown and Warner 1992; Liska and Yu 1992; Sampson and Laub 1993). Additionally, a label of dangerous may be assigned in response to a perceived economic or cultural threat that is not expressly political in character (e.g., Gusfield 1963; Auerhahn 1999b). The fact that the determination of dangerousness is a political process also influences the selection of targets. While it seems sensible to define dangerous crime as any behavior that is injurious or detrimental to society, the legal apparatus does not in fact treat all antisocial acts in the same way. Rather, the focus is on street crime, despite the fact that so-called white-collar or “corporate” crime is estimated to have a far greater negative impact on society overall (Geis 1968; Clinard 1990; Pearce and Tombs 1998).5 Floud and Young (1981) drive this point home by contemplating the irrationality of defining the behavior of individual criminals as gravely dangerous while simultaneously discounting the objectively greater risks that most people unthinkingly endure, such as the harms that result from industrial pollution and from exposure to toxic chemicals and other hazards in the workplace: There is little objectivity in perceptions of danger. It is a question of what people are prepared to put up with and why, and not simply of what is in some degree objectively damaging to them. Dangers are unacceptable risks: We measure or assess the probability and severity of some harm and call it a risk; but we speak of danger when we judge the
76
Incapacitation and Dangerousness
risk unacceptable and call for preventive measures. Risk is a matter of fact; danger is a matter of opinion. (Floud and Young 1981:4; see also Morris 1951; Lowrance 1976; Shah 1981) In chapter 6, I will articulate a measurement strategy that will be used to assess changes in the level of dangerousness in California’s prisons over time. Many policy researchers fall prey to the trap of tautology that inheres in the use of existing legal definitions to evaluate practical problems. In the present context, this takes the form of assuming that everyone in prison is dangerous— using as the sole decisional criterion the offender’s location in prison (e.g. Turner 1999). That is not analysis. If dangerousness is indeed a subjective construct, then we must ask ourselves whose subjective reality is to be privileged in determining what is dangerous. It is important to remember why we are considering dangerousness—in order to evaluate the efficacy of sentencing polices grounded in the logic of social defense. In light of this, it behooves us to take the views of the public that is presumably protected by such policies into account in formulating a definition of the dangerous offender.
DANGEROUS OFFENDERS When applying the label of dangerous to people, we must establish several things. First and foremost, what is the nature of the harm that we expect these people to do? Second, on what basis are we to determine the likelihood of an individual perpetrating that harm? Despite some variability in the precise definitions employed by different authors, there is a remarkable amount of consensus on what constitutes a dangerous offender. A few attributes emerge in nearly all contemporary discussions of dangerous offenders. It must be remembered that all definitions of dangerousness are inherently grounded in the logic of prediction. The attributes believed to contribute to the dangerousness of an individual are those that are believed to increase the likelihood of that individual perpetrating harm in the future. One characteristic that is universally mentioned in discussions of criminal dangerousness is violence (Morris 1951; National Council on Crime and Delinquency [NCCD] 1963, 1972; Wilkins 1985; Scott 1977; Rennie 1978; Floud and Young 1981; Conrad 1982; Moore et al. 1984; Austin 1986; Ancel 1987; Chaiken and Chaiken 1990; Pratt 1997). As one group of authors put it, “Violence is almost universally regarded as the hall-mark of dangerousness. Dangerous offenders are presumed to violent and violent offenders are presumed to be dangerous” (Floud and Young 1981:7). The focus on violence is consistent with the nature of dangerousness, particularly with respect to the dimensions of unpredictability, involuntary assumption of risks, and intent (McIntyre 1976;
Dangerousness
77
Shah 1981).6 Survey research concerning public perceptions of crime seriousness also supports the notion that the public places greater emphasis on the dangerousness of violent crime (Rossi, Simpson, and Miller 1985; Wolfgang et al. 1985; Carlson and Williams 1993). Repeated criminal behavior is also consistently identified as a flag for dangerousness. Various justifications have been advanced to support the idea of the repeat offender as a dangerous offender. Pratt (1997) identifies the quality of repetition that triggers the judgment of dangerousness as “ungovernability.” Repeated criminal behavior and a history of violent behavior as markers of dangerousness are logically grounded in anamnestic prediction—which is founded on the assumption that the future behavior of individuals is likely to be similar to the way they have behaved in the past (Morris and Miller 1985). Some scholars infer the motivations of the offender from the repetition of criminal behavior; this is evident in the terms “persistent recidivist,” “habitual offender”, and “incorrigible offender” (Morris 1951). All of these terms imply something about either the attitude of the offender toward crime (“habitual”), or the intended future conduct of the offender, as when persistence is inferred from past behavior. Similarly, “incorrigibility” implies a resistance to change. Mark A. Moore and his colleagues (1984) take this idea to extremes when they proclaim that Dangerous offenders are important not only because they are the most active offenders, but also because they are the guiltiest. They have committed criminal acts enough to have clearly revealed their character. After a fourth or fifth offense, the argument that the offender has values and character similar to other people in the society and was simply unlucky enough to find himself in tempting or provocative circumstances must yield to the view that the offender is much more willing than others to violate social rules. Such offenders have set themselves outside the moral order and exposed themselves to judgments of guilt. (Moore et al. 1984:30)7 These kinds of assumptions are not easily verifiable, and they also introduce an unnecessary degree of abstraction to the discussion. I shall refrain from construing dangerousness as something that lies in the offender’s state of mind. I believe this position to be indefensible in light of the nature of danger as a subjectively determined quality. Since dangerousness is socially defined, it properly speaks only to the attitudes of the potentially victimized community (or its representative) making the judgment that someone or something is dangerous. On this point I differ with those authors who would conceive of dangerousness as a propensity, present in all offenders to a greater or lesser degree (e.g., Gottfredson and Hirschi 1990). Rather, I take a much
78
Incapacitation and Dangerousness
more behaviorist approach to the measurement of dangerousness. While dangerousness itself is best thought of as a probabilistic and relative phenomenon, the operationalization of dangerousness must have a behavioral referent. In this sense, the term dangerousness is merely a convenient descriptor—it is used to characterize an intolerably high risk of the occurrence of a behavior that we believe to be unacceptable. The determination of dangerousness in an individual ultimately rests on a behavioral criterion; we cannot say that an individual is actually dangerous based on a calculated probability—however that estimate is arrived at. Indeed, using this strict behaviorist criteria, it is impossible to say that an individual is dangerous, in that such a judgment is implicitly predictive. Other frequently mentioned emblems of dangerousness rely on the logic of actuarial prediction—the estimation of a probability of criminal behavior based on an individual’s membership in a group. Two of these are gender (Allen 1987; Pratt 1997) and age (NCCD 1963, 1972; Kozol, Boucher, and Garolfolo 1972; Steadman and Cocozza 1974; Blumstein 1983; Farrington 1986). Men dominate all forms of criminal offending, particularly violence (National Research Council 1993; Maguire and Pastore 2001). Similarly, the inverse relationship between age and both the incidence and prevalence of criminal offending in the adult years has been called “the best-documented empirical regularity in criminology” (Nagin 1998:336; see also Farrington 1986).
DANGEROUS CLASSES AND DANGEROUS CRIMINALS Given that dangerousness is a social construction, the definitions of who and what is dangerous have changed with the times. Certain features of dangerousness persist in the selection of targets—perhaps none more prominently than that which Pratt (1997) has called “unknowability” (Pratt 1997).8 From the fourteenth through the seventeenth centuries, witches were almost universally considered to be dangerous throughout Europe and the English-speaking world (Erikson 1966; Rennie 1978). Witches were particularly “unknowable” due to the invisibility of their offense; it was impossible to prove or disprove by direct evidence, forcing the courts to rely on the testimony of those allegedly afflicted by the witches (Geis and Bunn 1997). This quality of incomprehensibility or unknowability may also explain the persistent association of dangerousness with the mentally ill. By the late nineteenth century, it was widely accepted among mental health workers in the United States that the mentally retarded were “moral imbeciles,” and possessed innate criminal tendencies (Deutsch 1949; Rafter 1997). While the term “moral imbecility” has since fallen out of fashion, the imputation of dangerousness to the mentally ill has persisted well into the twentieth century, despite the glaring lack of evi-
Dangerousness
79
dence that these individuals pose any greater threat than the general population (Steadman and Cocozza 1974; Thornberry and Jacoby 1979; Monahan 1981:115–118). Austin T. Turk (1982) contends that the label of “insanity” has been used at varying points throughout history to control individuals who pose a real or imagined threat to the political order (Turk 1982:52). Throughout the eighteenth and nineteenth centuries, it was considered common knowledge that “poverty is the mother of crime” (Rennie 1978; Pratt 1997). This belief manifested itself in the concept of the dangerous classes. Rennie cites an essay written on the subject in 1840 that highlights the clear distinction made between “wealthy, vicious idlers” and those who “join to vice the depravity of destitution . . . [one] does not become dangerous until he is without the means of existence or the desire to work” (quoted in Rennie 1978:3). In this case, the economic circumstances in which one found oneself could determine whether or not an individual was dangerous; situational, rather than behavioral, criteria determined dangerousness. This class-centered focus is revealed in the relatively greater abhorrence of property crimes, rather than violent offenses during this period (Rennie 1978; Pratt 1997). The idea of the dangerous classes fell out of favor toward the end of the nineteenth century. Rising up to take its place was the notion of the dangerous criminal. Several explanations have been advanced to account for the rise of an individualistic conception of danger. Pratt (1997) notes that in the Englishspeaking world toward the end of the nineteenth century, elite perceptions of internal threats to political stability lessened dramatically. Threats to the state came increasingly to be seen as emanating from external sources, rather than from a rebellious laboring class within. This perception arose from the increasing factionalization of the working class, which manifested itself in a number of other divisions such as those “between the deserving and undeserving poor; between the respectable and non-respectable; between labour aristocrats and unskilled workers” (Davies 1980:191). Also influencing this transition were advances in scholarly research and official record-keeping. The newly fashioned science of criminal statistics served to create a profile of the dangerous offender—in particular, the gathering of statistics provided evidence of the existence of the recidivist, or “habitual” criminal (Rennie 1978; Pratt 1997). Other scholarly influences on the changing conception of dangerousness were the burgeoning sciences of criminal anthropology and eugenics. The widespread acceptance of these ideas in America during the nineteenth century has been documented by Rafter (1997), who attributes at least some of their popularity to the status of science at the time, which she characterizes as “an admiration bordering on awe” (Rafter 1997:126). The dark figures of the “born criminal” and the “defective delinquent” were much more compatible with an individualistic—as opposed to class-based—view of the dangerous offender.
80
Incapacitation and Dangerousness
In the twentieth century, a renewed interest in the dangerous offender followed upon the heels of the demise of the rehabilitative paradigm. Several largescale research initiatives specifically devoted to the study of dangerous offenders were launched during this period, two in the United States and one in Great Britain.9 In the United States, the Lilly Endowment funded a group of researchers in Columbus, Ohio, in 1975. The broadest in scope of the research initiatives funded during this period, this project delved into a variety of topics including legal and historical aspects of dangerousness, the biological foundations of violence, juvenile and adult criminal careers, and criminal justice responses to dangerous offenders; the results of these investigations were reported in a number of books (Conrad and Dinitz 1977; Sleffel 1977; Hamparian et al. 1978; Rennie 1978; Van Dine, Conrad, and Dinitz 1979; Miller, Dinitz, and Conrad 1982). A few years later, the National Institute of Justice funded a group of researchers at Harvard University focusing primarily on issues relating to criminal incapacitation (Moore et al. 1984). In England, the Howard League for Penal Reform assembled a group of distinguished scholars, public officials, and criminal justice practitioners. Their investigation focused primarily on the practice of preventive detention and attendant ethical issues (Floud and Young 1981).
PREDICTION, CLASSIFICATION, AND DANGEROUSNESS As long as the notion of the dangerous offender has existed, so have attempts to identify and control him. Most of these attempts have been explicitly predictive; even those that are not, such as the variety of “classification” instruments used for inmate placement and programming in prisons, have an implicit predictive component, in that one objective of classification is usually the management/prevention of institutional misconduct on the part of the individuals being classified (Alexander 1986; Gottfredson 1987). Some of the technical limitations of our existing capabilities for predicting dangerousness were briefly discussed in chapter 4. Despite these difficulties, there has been a long history of the use of clinical prediction as justification for the incapacitation of individuals believed to be dangerous (Megargee 1976; Monahan 1981; Shah 1981). In the last thirty years, there has been a renewed interest on statistical methods for predicting dangerousness. Some of the most well-known examples of each are described below. Peter W. Greenwood (1983) has observed that clinical predictive strategies have historically dominated the “in” decision (e.g., sentencing and commitment), while objective methods of prediction (statistical and actuarial) have been more widely used to assist in making “out” decisions, such as parole release. John Monahan’s (1981) explanation for the massive overprediction of dangerousness in clinical settings may lend some insight into the preference for objective instruments in release decisions—the reification and impersonality of such instru-
Dangerousness
81
ments may serve to diffuse blame in the event of tragic errors in prediction (i.e., false negatives): If one overpredicts violence, the result is that individuals who will not be violent are institutionalized. This situation is not likely to have significant public ramifications for the individual responsible for the overprediction. But consider the other direction—underprediction. The correctional officer or mental health professional who predicts that a given individual will not commit a dangerous act is subject to severe unpleasantness should that act actually occur. (Monahan 1981:123)
THE FAILURES OF CLINICAL PREDICTION: BAXSTROM AND DIXON In 1966, the U.S. Supreme Court handed down a decision that would provide one of the first opportunities to study the validity of clinical predictions of dangerousness. Johnnie K. Baxstrom was originally convicted of assault in 1956 and sentenced to Attica. Baxstrom was diagnosed with mental illness while serving his prison sentence and sent to Dannemora, a facility for the criminally insane in upstate New York. In 1961, at the expiration of his sentence, he was held in Dannemora because he was deemed by facility staff to be “dangerously mentally ill.” Baxstrom petitioned that this proceeding violated his right to equal protection of the law, because the hearing process that resulted in his being detained at Dannemora was not subject to review by a jury, as it would have been in a civil commitment proceeding. The court concurred and ruled that Baxstrom was entitled to such review (Baxstrom v. Herold [1966]). It was subsequently determined that the Baxstrom decision applied to 966 other patients in the New York State system; rather than hold 967 hearings, the state simply transferred all of them to civil hospitals for the mentally ill (Steadman and Cocozza 1974). The transfer of these 967 patients to less secure facilities created the conditions for a “natural experiment”—an opportunity to assess the validity of the judgments of dangerousness that resulted in the detention of these individuals. Henry J. Steadman and Joseph J. Cocozza (1974) followed a sample of the Baxstrom patients (N 199) for four years. Their behavior in the civil hospitals and in the community after release was compared with an equivalent group (N 312) of patients who were transferred from the same state facilities for the criminally insane that had held the Baxstrom patients to civil hospitals before the Baxstrom decision (i.e., they were transferred because they were no longer considered to be dangerous). In the four-year follow-up, the researchers found that the Baxstrom patients were slightly more likely than the comparison group to be assaultive in the civil hospitals (15% vs. 6%), but point out that
82
Incapacitation and Dangerousness
more important that the relative differences between the two groups are the absolute findings on both groups. Eighty-five percent of the Baxstrom patients were not assaultive while in the hospital. . . . [T]he level of failure by the Baxstrom patients on the success criteria appear insufficient to support the psychiatric decision not to approve transfer. (Steadman and Cocozza 1974:106) Even more striking are the findings concerning the Baxstrom patients who were released to the community (N 98). Of these, only four patients (less than 2%) were returned to facilities for the criminally insane (Steadman and Cocozza 1974:103). Twenty patients were rearrested, but the vast majority of the offenses were nonviolent or minor offenses (Steadman and Cocozza 1994:139). Steadman and Cocozza combined all the cases where a released patient exhibited violent behavior (resulting in either recommitment or arrest), and found that only fourteen of the released patients (15%) ever behaved in a way that could have been considered dangerous during the follow-up period (Steadman and Cocozza 1974:151). It is important to remember that every one of these individuals was initially deemed to be so dangerous as to justify his indefinite confinement in a facility for the criminally insane. Steadman and Cocozza’s study of the Baxstrom patients sent shock waves through the clinical community, calling the entire enterprise of clinical prediction into question (Monahan 1981). Five years after the Baxstrom decision, a similar case came before the federal court. This class action suit was filed by Donald Dixon and six other named plaintiffs. All were inmates of Fairview State Hospital, a facility for the criminally insane in Pennsylvania. The suit challenged the constitutionality of the Pennsylvania statute that allowed for the indefinite confinement of an individual beyond the expiration of a criminal sentence (Thornberry and Jacoby 1979). In a ruling similar to the Baxstrom decision, the Dixon court found that the Pennsylvania procedure violated due process protections afforded to civilly committed patients (Dixon et al. v. Attorney General of the Commonwealth of Pennsylvania [1971]). As a result of the Dixon decision, 586 inmates of Fairview State Hospital were transferred to civil hospitals. Terence P. Thornberry and Joseph E. Jacoby performed a follow-up study that was nearly identical in method to that of Steadman and Cocozza (1974). The findings and conclusions of Thornberry and Jacoby were also nearly identical to those of Steadman and Cocozza. A mere 14.5 percent of the Dixon patients—all of whom were determined to be so dangerous as to preclude even their transfer to a less secure facility—were found to have committed either a violent criminal offense or other violent act that resulted in rehospitalization. Indeed, Thornberry and Jacoby quote Steadman and Cocozza in summing up their own findings:
Dangerousness
83
If we were to attempt to use this information for statistically predicting dangerousness our best strategy would be to assume that none of the patients were dangerous. In this case, we would be wrong in 14 of the 98 cases. Any other method would increase our error. (Steadman and Cocozza 1974:151; emphasis in original; see also Wenk, Robison, and Smith 1972)
DEATH AND DANGER IN TEXAS The results of another such “natural experiment” were reported by Marquart et al. (1989). In 1972, the U.S. Supreme Court ruled that the death penalty as administered in the United States constituted cruel and unusual punishment and was therefore unconstitutional (Furman v. Georgia [1972]). After the Furman decision, states restructured their death penalty statutes in order to bring them into compliance with constitutional requirements. In Texas, the reformulated statute provided for a bifurcated trial procedure for persons accused of capital murder. After a finding of guilt, capital murder defendants would then be granted a punishment hearing. During the punishment hearing, it fell to the jury to determine whether the defendant would receive a sentence of life imprisonment or death. This determination centered largely on the jury’s answers to three questions. All three had to be answered in the affirmative for a defendant to receive a death sentence; if any question was answered in the negative, the defendant would be automatically sentenced to life imprisonment. The first establishes criminal intent; the third establishes that the force used by the defendant cannot be deemed justified, given the circumstances (Marquart et al. 1989:450). The authors found that the first and third questions were almost always answered in the affirmative, which meant that the second question was the one that distinguished between those defendants receiving a life sentence and those receiving death (Marquart et al. 1989:451). This question entailed an explicit prediction of the defendant’s future dangerousness: “whether there was a probability that the defendant would commit criminal acts of violence that would constitute a continuing threat to society” (Texas Criminal Procedural Code [1985], cited in Marquart et al. 1989:450).10 Marquart et al. (1989) examined the institutional records of 92 inmates sentenced to death under the revised statute—all of whom were believed to “constitute a continuing threat to society”—who subsequently had their sentences commuted to life imprisonment. 11 The authors used as a comparison group of 107 defendants who were sentenced to life imprisonment under the same statute—because juries did not believe them to be dangerous. The
84
Incapacitation and Dangerousness
analysis showed that the two groups differed very little with respect to their subsequent behavior; the death-sentenced inmates had, on average, lower rates of serious institutional misconduct than those not deemed to be dangerous by jurors, and both groups demonstrated equivalently low rates of rearrest upon release (Marquart, Ekland-Olson, and Sorenson 1989).
OBJECTIVE INSTRUMENTS FOR PREDICTION OF DANGEROUSNESS
THE
The use of objective prediction/classification instruments in American criminal justice dates back to 1928, when Ernest W. Burgess created an instrument for predicting success on parole release. Burgess analyzed the postrelease behavior of 3,000 parolees in Illinois and constructed contingency tables of their outcomes with respect to twenty-two factors, including criminal history, family history, social factors (such as marital status and employment), and adjudication information (such as sentence length, and whether or not sentence was part of a plea bargain) (Bruce et al. 1928:205–249). Burgess then constructed a simple predictive scale. For each factor, one point was awarded if the individual fell into the category with a higher than average success rate on parole; the points were then summed to provide a single score. The higher the individual’s score, the more favorable his prognosis on parole. In an early validation study, Hakeem (1948) analyzed 1,108 parolees and found that the predictions made using the Burgess method demonstrated “remarkable accuracy” (Hakeem 1948:386). The use of the Burgess method has persisted in parole administration. A predictive instrument based on this method, the Salient Factor Score (SFS 81), is still used today by the U.S. Parole Commission (Hoffman 1994). Introduced in 1972, the Salient Factor Score is a simple additive construct consisting of six items; these include a variety of indicators of criminal and incarceration history and a single item regarding a history of drug dependence. Individuals receive one or more points if they fall into the “favorable” category on a particular item. Possible scores range from 0 to 10, with a score of 8–10 representing a “very good” prognosis and 0–3 representing a “poor” one (Hoffman 1994). The Salient Factor Score has been shown to be both valid and remarkably stable over time, even when tested against more stringent outcome criteria (in the form of longer follow-up periods) and on a different population than that for which it was originally intended (Hoffman 1982; Hoffman and Beck 1985; Hoffman 1994). The Greenwood/Abrahamse selective incapacitation proposal (1982) discussed in the previous chapter utilized a similar Burgess-style additive index. The authors claimed that the seven-item scale could successfully distinguish between dangerous and nondangerous offenders (called “high-and low-rate” in
Dangerousness
85
the proposal), and recommended that predictions about the future behavior of offenders derived from the scale be used to aid in making sentencing decisions. The items used in scale construction included a variety of indicators of juvenile and adult criminal history, drug use, and employment history. As the tables in chapter 4 show, these claims are insupportable, in that the instrument fails to identify high-rate offenders with an accuracy greater than would be expected due to chance alone (see Auerhahn 1999a for an expanded discussion of the limitations of this particular proposal). In the context of parole prediction, the criterion variable (i.e., “dangerousness”) is usually defined as the commission of any new crime while on parole release (Hoffman 1982; Hoffman and Beck 1985; Gottfredson and Gottfredson 1994; Hoffman 1994). The other major use of objective prediction instruments is in inmate classification instruments, which generally seek to predict institutional misconduct (Alexander 1986; Buchanan, Whitlow, and Austin 1986; Kane 1986; Fernandez and Neiman 1998). The use of these types of instruments increased in the 1980s, as a result of the combined effects of the widely publicized failures of the clinical prediction techniques that had traditionally been used for determining placement and program needs, and of the increased intervention into prison operations by the courts, which came about as the result of a number of inmate lawsuits challenging existing classification schemes based on clinical prediction (Austin 1983). Research results evaluating the success of prison classification instruments are mixed. Some researchers find that these instruments do no better than chance at predicting inmate dangerousness (Proctor 1994; Fernandez and Neiman 1998), while others report more favorable results (Buchanan et al. 1986; Austin 1993; Coulson et al. 1996). However, even those instruments that successfully distinguish between low- and high-risk inmates are not free of problems. Austin (1983) used simulation analysis to examine the operation of several of these models and found that despite differences in the construction of the models (the number of factors included ranged from 6 to 24), all produced remarkably similar classification distributions. Perhaps most striking was his finding that all the models examined were principally driven by two generic factors: the inmate’s current offense and the inmate’s previous criminal history . . . . social factors (age, education, etc.) exert virtually no influence on aggregate scores. Most significantly, information on the inmate’s prior institutional behavior, the behavior that classification models strive to predict, are of little importance. (Austin 1983:569) Austin (1986) reported similar findings for an assessment of California’s inmate classification instrument. He found that a single variable, sentence length,
86
Incapacitation and Dangerousness
accounted for 79 percent of the variance in classification scores; the other twenty-three items in the instrument exerted virtually no influence on an inmate’s classification (Austin 1986:310–313). In a more recent evaluation of the California inmate classification system, Fernandez and Neiman (1998) reported a similar finding; they also found that the instrument failed to accurately predict serious institutional misconduct. In sum, the objective prediction instruments appear to be better suited for some purposes than others. Such instruments are clearly more successful in the context of parole prediction than in predicting dangerous behavior in (or out of ) prisons. The reason for this is simple. In parole prediction, the goal is the identification of individuals who are the least dangerous, while the goal of inmate classification is to identify those individuals who pose the greatest amount of risk. Simply put, the two uses of prediction are geared toward identifying different ends of the range of behavior. One predicts danger; the other predicts safety. Parole prediction, in seeking to identify the nondangerous, has the advantage of a higher base rate, which increases the chance of a successful prediction.
CONCLUSION Danger has proven to be an elusive concept in criminology. The foregoing discussion makes clear that dangerousness has much more to do with the subjective judgments of other than with the actual behavior of those who receive the label. Additionally, attempts to identify and control dangerous offenders have enjoyed only limited success. However, this does not necessarily mean that dangerousness is not a useful concept. Many useful lessons have been learned from attempts to predict dangerousness in individuals. The following chapter offers a strategy that draws upon these lessons in order to evaluate the success of selective incapacitation strategies in California.
Chapter 6 Assessing the Level of Dangerousness in the Criminal Justice System
The protection of the public by the incapacitation of dangerous offenders has long been regarded as a legitimate aim of the California criminal justice system. In recent years, sentencing policy in California has reflected a growing interest in this objective, demonstrated most prominently in recent reforms such as Truth in Sentencing and the state’s Three Strikes habitual offender statute. However, early investigations into the effects of these laws indicate that these objectives may not be being achieved in practice. For example, numerous studies report that the overwhelming majority of Three Strikes defendants are convicted of nonviolent offenses (LAO 1996, 1999; Austin 1998), while others emphasize the relatively older age of Three Strikes defendants, citing the futility of incarcerating offenders for lengthy terms, when they are likely to “age out” of criminal behavior of their own accord (Zimbardo 1996; Austin 1998). Other research indicates that the Three Strikes law has profoundly altered the composition of the jail population in California, with Three Strikes defendants pushing out other pretrial and sentenced jail inmates (Turner 1999; LAO 1999). While it is too early to definitively report on the impact of the Three Strikes law, previous research hints at the likely systemic consequences. Bales and Dees (1992) examined the impacts of mandatory minimum sentences in Florida. During the 1980s, the Florida legislature enacted legislation that targeted certain offenses for mandatory minimum sentences. The majority of inmates sentenced under such laws during the period of the analysis were drug and otherwise nonviolent offenders (Bales and Dees 1992:320). During the same period, Florida’s prisons were placed under court-ordered capacity limits requiring corrections officials to balance new admissions with releases. These researchers found that mandatory minimums reduced the size of the pool of 87
88
Incapacitation and Dangerousness
inmates eligible for release, which had the net effect of reducing the average sentence length for all nonmandatorily sentenced inmates, including violent offenders (Bales and Dees 1992). Given the state of persistent crowding in California’s prisons, it seems reasonable to anticipate a similar outcome in California. Over two-thirds of new prison commitments under the Three Strikes law are for nonviolent offenses (LAO 1999). The strategy of retrospectively identifying habitual offenders embodied in the “Three Strikes and You’re Out” statute results in a population of Three Strikes prison admissions with an average age of thirty-six—ten years older than the average offender sentenced to prison (Austin et al. 1999). The likelihood that older offenders will “age out” of criminal offending combined with the level of crowding in California’s prisons suggests that the impact of sentencing reforms intended to incapacitate dangerous offenders may well be a net reduction in the level of dangerousness in the incarcerated population. Although the state’s prisons are not currently operating under a court-ordered population cap, there are basic physical constraints on institutional capacity. Since California prison facilities are already operating at approximately 200 percent capacity (Bureau of Justice Statistics 2001), it seems that this limit will soon be reached, necessitating the release of substantial numbers of prisoners to counter rising incarceration rates. California has the largest correctional system in the United States, holding over 163,000 adult inmates in state prisons and nearly 80,000 in local jail facilities (Bureau of Justice Statistics 2001; California Department of Justice 2002). The state also supervises over 100,000 adults on parole and over 300,000 adults on probation (California Department of Corrections 2002; California Department of Justice 2002). Although the number of persons under all forms of correctional supervision has grown at an alarming rate in the last two decades, the greatest amount of concern has been expressed about the growth of prison populations. Prisons are by far the most expensive form of correctional supervision. In fiscal year 1998–1999, California spent over $4 billion to operate the state prison system (California Department of Justice 2002). The growth of the prison population has far outpaced population growth in the state. In 1976, there were 81 prisoners per 100,000 state residents; in 1997, there were 475 prisoners per 100,000 (Gilliard and Beck 1998; Maguire and Pastore 2001). California’s Three Strikes law is likely to exacerbate this situation. Since the law’s passage in 1994, over 40,000 inmates have been sent to prison under the provisions of the statute. This stands in stark contrast to other states with similar laws. For example, Washington has sentenced fewer than 125 offenders under its 1993 Three Strikes statute; Nevada has sentenced fewer than 200 (Litvan 1998).1 The empirical objective of this study is to assess the efficacy of California’s sentencing structures in terms of the goal of protecting the public by
Assessing the Level of Dangerousness in the Criminal Justice System
89
incapacitating the dangerous. The remainder of this chapter will be devoted to laying out the strategy that will be used to accomplish this end.
DANGEROUSNESS AS AN EVALUATIVE CONSTRUCT Chapter 5 demonstrated that there has been, historically, a great deal of interest in the prediction of dangerousness. That discussion also showed that a great deal of evidence exists that speaks to the limited success of these endeavors. Sentencing innovations such as Three Strikes and Truth in Sentencing aim to select the most dangerous offenders for lengthy incarceration, thereby isolating them from society. However as just discussed, due to structural constraints, the net effect of these laws may be a reduction in the aggregate level of dangerousness in the prison population, as more dangerous offenders who are not subject to mandatory minimums are released to make room for Three Strikes and for other mandatorily sentenced offenders. The strategy that will be employed in this research to assess dangerousness diverges from previous research in a number of ways. First, and most importantly, the analysis does not attempt to predict—or even assess—the dangerousness of individual offenders. Rather, dangerousness is here conceived as a stochastic property of aggregates, rather than one of individuals.2 This conceptualization is informed by both the failures and lessons of prediction. As was shown in chapter 5, an examination of the prediction literature leads to the unmistakable conclusion that predictions of individual dangerousness cannot be made with any reasonable level of accuracy; prior attempts to prospectively identify dangerous offenders have been characterized by false-positive rates greater than 50 percent (e.g., Steadman and Cocozza 1974; Thornberry and Jacoby 1979; Monahan 1981; Greenwood and Abrahamse 1982; Auerhahn 1999a). However, there are valuable lessons contained in this research. Most notably, we have learned from these failed attempts to predict that although dangerousness cannot be accurately predicted in individuals, there are several reliable correlates of dangerousness. What this means is that while we cannot know whether a particular individual—for instance, a nineteen-year-old male with a lengthy history of violent behavior—is likely to continue to be dangerous, we do know that on average, nineteen-year-old males who exhibit a pattern of assaultive behavior tend to be overrepresented in the universe of dangerous persons. Put another way, while most young people are not dangerous criminals, most dangerous criminals are young (see Chaiken and Chaiken 1983:28 on this point). Although we are not able to accurately predict which individual offenders are likely to be dangerous, we can retrospectively identify the correlates of dangerousness in large numbers of individuals, and we can also say with some confidence that some percentage of a population of offenders is likely to
90
Incapacitation and Dangerousness
be dangerous. While we have very little confidence in the assertion that “individual A is dangerous,” it is actually quite plausible to comment on the composition of a population and conclude that “ it is likely that population A contains more dangerous individuals than does population B.” In this analysis, dangerousness is used as an evaluative construct to estimate the level of dangerousness in correctional populations.3 Populations are, of course, comprised of individuals. The modeling strategy employed here consists of measuring certain characteristics of individuals known to be correlated with dangerousness. The level of dangerousness in a population is thus defined as the mean level of dangerousness in individuals comprising the population. If get-tough sentencing policies based on the objective of selective incapacitation have been successful, we should see an increasing level of dangerousness in prison populations over time, with corresponding decreases in the dangerousness level of other criminal justice populations. If the analysis indicates a decrease in the level of dangerousness in the prison population over time, this would indicate that sentencing policies aiming to protect the public by incapacitating dangerous offenders have not been successful in achieving this objective. This approach allows for many instructive comparisons. For instance, it is not only useful to examine trends within particular system populations over time (e.g., prison), but we can also learn a great deal about the systemic effects of sentencing reform by examining trends in the level of aggregate dangerousness in different populations over time. For example, we may find that the level of dangerousness in the prison population is declining over time, while at the same time the level of dangerousness in the parole population has been increasing. Such a finding would tend to support the existence of a cumulative effect similar to that discovered by Bales and Dees (1992) in the Florida system.
OPERATIONALIZING DANGEROUSNESS In order to make dangerousness into a useful operational construct, it is necessary to consider the multiple bases upon which one might base a determination about the dangerousness of an offender. Most empirical analyses of “dangerous offenders” have failed to do this, with most researchers operationalizing dangerousness as either violence (e.g., Steadman and Cocozza 1974; Thornberry and Jacoby 1979; Monahan 1981) or persistence (e.g., Greenwood and Abrahamse 1982), but not both. Dangerousness has multiple indicators, and multiple manifestations (Menzies et al. 1985). For instance, Chaiken and Chaiken (1990) make a distinction between high-rate and violent offenders—but consider both to be dangerous. Similarly, Scott (1977) points out that “individually non-dangerous offences, if repeated sufficiently often, achieve dangerousness by their threat to the rule of law” (Scott 1977:128). Dangerousness is both multidimensional and relative. Given the probabilistic nature of dangerousness, it
Assessing the Level of Dangerousness in the Criminal Justice System
91
TABLE 6.1 Coding Scheme for Dangerousness Measure Characteristic
Score
Age 18–24 25–34 35–44 45–54 55 and older
4 3 2 1 0
Male Female
1 0
2 1 none
2 1 0
1 or more none
1 0
violent property drug
3 2 1
Gender
Prior felony convictions
Prior violent convictions
Current conviction offense
Minimum score: 1 (55 year-old woman, no priors, drug offense) Maximum score: 11 (18–24 year-old male, 2 or more priors, violent prior[s], violent offense)
is much more sensible to consider dangerousness as a continuum rather than as a dichotomy. Consider the following hypothetical offenders: Offender A, convicted of forcible rape, and Offender B, who has a string of convictions for residential burglary and car theft. Both of these offenders would be considered dangerous by just about anyone, and a legitimate target for selective incapacitation. But now consider Offender C—with a long criminal history that also includes violence. It seems reasonable to suppose that we would consider this last offender to be even more dangerous than either of the first two. The dangerousness construct employed in this research takes this multidimensionality into account. It consists of a relatively simple, Burgess-style index comprising multiple indicators of the correlates of what is universally considered dangerous criminal behavior. The dangerousness construct as measured in individuals takes the form of an additive index, comprised of the following variables (the coding scheme is given in table 6.1):
92
Incapacitation and Dangerousness
• age (in groups 18–24, 25–34, 35–44, 45–54, and 55 and older) • gender (male/female) • number of prior felony convictions • presence or absence of violent prior felony convictions • current conviction offense The dangerousness measure itself is a fairly simple construct. It is similar in structure to the Legal Dangerousness Scale used by Steadman and Cocozza (1974), which consisted of information on juvenile criminal history, number of prior incarcerations, presence or absence of violent convictions, and severity of the conviction offense (Steadman and Cocozza 1974:107). The researchers found that the combination of the offenders’ age and his score on this simple four-item scale successfully distinguished between patients who were rearrested and/or rehospitalized and those who were not (Steadman and Cocozza 1974:146–147).
DEMOGRAPHIC VARIABLES The age coding structure reflects the sharply descending probability of criminal behavior as offenders age. Rates of criminal offending peak in the teenage years, and decline sharply thereafter (Farrington 1986; but see also Figlio 1996). While the curves for violent and property offenders differ slightly in shape, the form of the relationship between age and offending closely approximates linearity. This is of the utmost importance in considering the impact of policies whose goal is to incapacitate dangerous offenders: “From the standpoint of incapacitation, the longer the time served, the more likely it is that the individual would have terminated his criminal activity even if he were not in prison. In this sense additional prison time is ‘wasted’” (Blumstein 1983:245). The primary reason for including gender in the dangerousness construct is to examine the systemic impact, if any, of the pattern of sharply rising incarceration rates for women. Nationwide, between 1986 and 1991, women’s incarceration rates increased 75%, while the incarceration rate for men increased 53% (Greenfield and Minor-Harper 1991). Female offenders tend to have less extensive criminal histories and are less likely to be convicted of violent offenses (Bureau of Justice Statistics 199l; Bloom, Chesney-Lind, and Owen 1994). Men are also considered more dangerous than women in the coding scheme due to their dominance in all measures of criminal offending. Males account for approximately 80% of all arrests, and an even higher proportion of arrests for violent crimes (Maguire and Pastore 2001). Some researchers might argue for the inclusion of race in a measure of criminal dangerousness, based primarily on the disproportionate representation of nonwhites in official criminal statistics (e.g., Blumstein 1982; Blumstein and
Assessing the Level of Dangerousness in the Criminal Justice System
93
Graddy 1982). There are several reasons for excluding race in the measurement of dangerousness used in this study. One such reason is that arrest data may reflect a good deal more than simply the criminal activity of the offender—insofar as group differences in the probability of arrest may be independent of criminal involvement to some degree (Black 1970; Smith 1986). The weight of this argument is somewhat diminished by analyses of victimization survey data that support the notion of disproportionate minority involvement in criminal behavior (Hindelang 1978; Langan 1985). However, as far as dangerousness is concerned, while Alfred Blumstein and Elizabeth Graddy (1982) found racial differences in criminal involvement as measured by prevalence, the incidence of criminal behavior—the probability of rearrest given a first arrest, or the dimension of persistence—was strikingly similar across racial groups (Blumstein and Graddy 1982:288), and is thus not a uniquely useful indicator of dangerousness. Furthermore, race is confounded with a number of other factors that are likely to be related to differential criminal involvement, most notably socioeconomic status (Wilson 1978), thus hindering our ability to impute the unique effects of race on criminal dangerousness.4
CRIMINAL HISTORY VARIABLES The use of criminal history information in the dangerousness construct reflects the fact that prior criminal behavior is the best predictor of future criminal behavior. David P. Farrington (1979) reviewed a number of longitudinal studies and found a pattern of “steeply rising probabilities” of reoffending after each subsequent criminal conviction (Farrington 1979:302). The dangerousness construct counts up to two prior felony convictions. There are several reasons for this. The first is that this corresponds to the requirements of the Three Strikes statute, allowing for the estimation of the impacts of this law on the dangerousness of criminal justice populations. Information on convictions (rather than arrests, or some other measure of criminal offending) is used due to the greater reliability of this type of information, as well as its greater accessibility in official records for purposes of model validation.5 The ranking of the current conviction offense according to seriousness is consistent with the relativistic and continuous nature of dangerousness. Conceiving of dangerousness as a continuous variable allows us to prioritize among different degrees of dangerous offenders—something that is of the utmost importance in the face of structural constraints on incarceration resources. Thus, considering violent offenses as relatively more dangerous than property offenses does not in any way discount the dangers posed to society by property offenders. Rather, the dangerousness construct merely employs a weighting scheme based on the priority schedule that most individuals would assign to limited resources
94
Incapacitation and Dangerousness
for restraining offenders. Few people would argue that they would rather see scarce prison space allocated to a burglar than to a violent rapist or murderer (although most would like to see both offenders incarcerated). Similarly, the presence of a violent criminal conviction in an offenders’ prior record receives greater weight in the dangerousness construct than does a nonviolent criminal history. There is some evidence that crimes of violence are more dreaded than are other crimes without an interpersonal aspect (McIntyre 1976; Warr and Stafford 1983; Warr 1984).6 While violent crimes are relatively rare in comparison to property and drug offenses, I am in complete agreement with those authors who submit that fear of crime is a social fact with tangible social consequences—and that whether that fear is “rational’ with respect to actual victimization risk is not necessarily relevant in all situations (Warr 1984; Ferraro 1995). Danger is about the subjective determination of some risks as unacceptable. In evaluating a policy whose stated objectives include protection of the public, it seems only reasonable that we take into consideration the things from which the public would most like to be protected.7 As Zimring and Hawkins (1997) note, “by longstanding habit, Americans use the terms ‘crime’ and ‘violence’ interchangeably” (Zimring and Hawkins 1997:3). Violent offenses are here defined broadly, as all offenses involving a proximate victim, regardless of whether harm comes to that victim or not. This definition includes all offenses involving interpersonal contact, including rape, homicide, robbery, and assault. Property crimes include offenses involving theft, larceny, and fraud, such as auto theft and burglary. Drug offenses are self-explanatory, and include all types of drug crime, such as possession, distribution, and manufacture. Based on the findings reported in the criminal seriousness literature, violent offenses will contribute the most to the dangerousness score, followed by property offenses, with drug offenses considered the least dangerous. While there is evidence that most people rank certain drug offenses (e.g., running a narcotics distribution ring) as more serious than many violent crimes (Wolfgang et al. 1985), persons found guilty of such offenses are a relative rarity in the criminal justice system. A recent analysis of drug offenders in federal prisons found that 36% of all drug offenders could be classified as “low-rate” using the rather stringent criteria of no prior incarcerations, no history of violence and “no sophisticated criminal activity” (U.S. Department of Justice 1994).8 An analysis of all federal prisoners released in 1987 revealed that lowlevel drug offenders had a recidivism rate of 20%, as compared to 41% for all released prisoners (U.S. Department of Justice 1994).
SYSTEM SIMULATION The strategy for evaluating the selective success of sentencing policy reform over time makes use of a relatively underutilized technique in sociology and crimi-
Assessing the Level of Dangerousness in the Criminal Justice System
95
nology: continuous-state continuous-time dynamic systems modeling. Hanneman and Patrick (1997) go so far as to describe the use of dynamic modeling in the social sciences as “exotic” (Hanneman and Patrick 1997:1). There is little innovation in quantitative analytic techniques in sociology—static analyses using multiple linear regression analysis and extensions thereof, such as logistic regression, path analysis, and covariance structure analysis dominate the field (Patrick 1991). Similarly, it is rare for sociologists to model the systemic properties of social phenomena—most empirical research is predicated on simple, unidirectional causal models, which do not permit the analysis of complex, multidimensional processes, nor do they afford any insight into the process by which structures change (Hanneman 1988). This is a serious limitation of social science, as most social phenomena are not in fact unidirectional chains of events, but rather continuous and interrelated processes in which outcomes later in the cycle influence the processes that give rise to those outcomes over time. Take this simple example: a young couple decides, early in their marriage, to have five children. A number of factors have likely influenced this decision in a causal sense (e.g., education, income, religiosity, and level of “traditional” value orientation). However, it is conceivable that this couple may, after having had two children, rethink the original decision. It would be difficult to argue that having the first two children had no impact on the couple’s reconsideration of the earlier decision. In systems modeling terms, this kind of process is characterized as feedback (Forrester 1969b). Most social processes are characterized, to a greater or lesser degree, by some type of feedback mechanism. An example of a process in the California criminal justice system that may be influenced by feedback can be seen in the problem of parole violation. There has been some consternation in recent years over rates of parole violation, with over half of released offenders being returned to prison (California Department of Corrections 2002; see also Costello, Garnett, and Schiraldi 1991). While many of these offenders are returned for technical violations of the conditions of parole, some are recommitted for new offenses. Some of these new offenses may represent the systemic effects of the implementation of politically popular sentencing reforms. In order to accommodate offenders under various mandatory sentencing provisions, other offenders who may be more dangerous but who are not subject to the provisions of mandatory sentencing schemes are released, upon which they commit further criminal acts and are returned to prison, further exacerbating the strain on system resources. While it has become commonplace for criminologists to speak of “the criminal justice system,” most research tends to focus on static analyses of single components of the system (e.g., jails and prisons), rather than conceiving of the entire system as a system (important exceptions include Blumstein and Larson 1969; Cassidy and Turner 1978; Cassidy, Peters, and Turner 1981; Cassidy 1985; Ohlin and Remington 1993). However, the impacts of legislative changes that are intended to act upon one component of the system may
96
Incapacitation and Dangerousness
have unintended systemwide consequences. Analyzing the impacts of legislative changes to sentencing structures from a systems perspective can provide important and useful insights into the unintended consequences that may result from these changes. At the simplest level, systems are comprised of a number of distinct parts that are interrelated in some way (Hanneman 1988; Teller 1992; Dorny 1993). In this analysis, the “parts” are criminal justice population-states9 such as jail, prison, and parole. They are connected by the structure of the adjudication process, as depicted in figure 6.1.10 The main structural components of a dynamic system are levels and rates, or stocks and flows (Forrester 1969b; Hanneman 1988). Levels or stocks can also be thought of as accumulators, in which whatever is contained in the stock (e.g., persons, emotional energy, etc.) is conserved over time (High Performance Systems 1997). The rates or flows represent the processes by which conserved quantities move from one level to another—in this case, the probability of an individual moving from one state of the system (a stock, such as jail or prison) to another in a given space of time. The “fundamental process” of system change is the integration of these movements over time (Forrester 1969b:21). The focus on integration represents an epistemological departure from the
FIGURE 6.1 Structural Model of the California Criminal Justice System
Assessing the Level of Dangerousness in the Criminal Justice System
97
logic of conventional statistical and time-serial methodological applications in that these traditional methodologies tend to consider (with varying degrees of qualification) observations to be independent of one another—and if not independent, at least as distinct from one another. Even methodological approaches that concede the nonindependence of observations (e.g., panel analysis, ARIMA modeling) consider observations to be influenced by proximate observations, but do not consider that different observations may be comprised of the same components, as is likely to be the case when comparing the state prison population at two consecutive yearly intervals. The modeling strategy used to evaluate the efficacy of sentencing reform with respect to selective incapacitation consists of simulating the movement of individuals through these population states, in an attempt to reproduce the actual historical composition of these populations that has been observed for the period under study (1970–1998). It is a widely acknowledged fact that offenders move from one criminal justice system state to another at different rates. A number of factors that influence transition probabilities within the system are legally relevant, such as prior criminal record or type of conviction offense. Others are not necessarily legally relevant, but are relevant vis-à-vis the dangerousness construct, such as gender and age (Irwin 1985). In addition to these, other variables which are neither legally relevant nor salient to the dangerousness construct have been shown to influence the rate at which offenders transition from one system state to another. The most well-documented of these is the race of offenders (Bridges and Crutchfield 1988; Tonry 1994, 1995; Crawford, Chiricos, and Kleck 1998; Gorton and Boies 1999). In order to accurately reproduce system dynamics, the simulation modeling process must take into account all variables that are know to significantly influence the transition probabilities, including race.11 This necessitates the construction of a number of structurally equivalent models (e.g., the model depicted in figure 6.1) for individuals with similar groupings of attributes. This results in four hundred fifty separate models which must be estimated simultaneously. The results of these simulations are then combined to reconstruct the model of California criminal justice system in its entirety (Appendix A describes the modeling process in greater detail). The concept of emergence is central to the analysis of dynamic systems (Patrick 1991; Dorny 1993; Gilbert 1995; Mihata 1997). Emergence is frequently understood as the development of structures and patterns of action arising out of the actions and interactions of individual agents (Schelling 1978; Epstein and Axtell 1996). Mihata (1997) defines emergence somewhat more broadly: “emergence is characterized by a nonlinear mode of organization that can generate nonobvious or surprising consequences” (Mihata 1997:32; see also Teller 1992). In the present analysis, emergence is understood in terms of systemic structural consequences that occur as a result of actions intended to have an impact on one system element, but that influence the constitution of other
98
Incapacitation and Dangerousness
system elements in an unintended fashion. An example of this type of emergent condition was alluded to earlier; there is evidence that California’s Three Strikes law, which is intended to influence the composition of prison populations, has had serious indirect and unintended effects on the composition of jail populations (Turner 1999). The ability to understand and observe the processes that give rise to these emergent properties of systems afforded by simulation analysis is one of the greatest strengths of this methodology. Following Crutchfield (1992), Byrne (1997) draws a distinction between the “engineering” and the “scientific” approach to simulation modeling. The engineering approach is applied—the primary consideration is the efficacy of the model at representing the empirical reality of interest. The scientific approach is much more concerned with understanding the mechanisms that give rise to that reality, as when simulation modeling is used in the construction and testing of theories (e.g., Jacobsen and Bronson 1985; Hanneman 1988; Patrick 1991; Axelrod 1995; Hanneman, Collins, and Mordt 1995). While the present endeavor falls somewhere in between these two extremes, my approach to modeling the California criminal justice system is much more compatible with the engineer’s approach. This analysis is concerned with specific outcomes (the composition of the prison and other correctional populations over time) that come about as a result of a general process (paradigm change) in a particular time and place (California in the late twentieth century). The purpose of this type of simulation exercise is summed up by Byrne (1997): “If we are dealing with a world characterized by emergent properties then what we want to be able to describe is the way in which those properties emerged. This is not a process of analysis but it is a process of historical account” (Byrne 1997:5). The present examination of the systemic consequences of sentencing reform in California is less a causal analysis than a functional one. This is not an exercise in theory construction or hypothesis testing, but rather an exercise in policy evaluation and development. Simulation modeling has been described as an attempt to “describe through experimentation” (Mihata 1997:35). This is particularly applicable here; the first goal of the simulation is not prediction, but rather an attempt to explain how a particular set of circumstances came about. Static data on the composition of California criminal justice system populations are available for the period that will be simulated;12 however, these data do not tell us anything about the emergent properties of the system that created the realities they represent. Dynamic systems simulation modeling is uniquely suited to addressing the question of the efficacy of selective incapacitation vis-à-vis dangerous offenders. Mihata (1997) has observed that “computer simulation makes possible new kinds of knowledge about complex systems—and possibly, new explanations for emergents that we have been able, to this point, only to intuit” (Mihata 1997:36). The problem under study here illustrates this point admirably. Sev-
Assessing the Level of Dangerousness in the Criminal Justice System
99
eral authors have suggested that incarcerated populations may be becoming less dangerous over time as a result of the unintended consequences of sentencing reform (Costello, Garnett, and Shiraldi 1991; Bales and Dees 1992; U.S. Department of Justice 1994; Canela-Cacho et al. 1997; Auerhahn 1999a), but conventional statistical methods are not equal to the task of investigating this possibility. Due to the complexity of the equation system needed to address this question, mathematical approaches are foiled by the identification problem. Simply put, the large number of transition probabilities between system states needed to represent the dynamic system results in more unknown values than known values in the equation system. In this situation, no single set of definitive parameter estimates can be obtained. Accurately representing the dynamics of the California criminal justice system requires a degree of complexity that renders direct solution infeasible.13 The first goal of the simulation is to accurately mimic the structural features of the system that have resulted in the current composition of the California criminal justice system. This leads to the second component of the analysis, which I call predictive evaluation. Predictive evaluation is a form of what Conte and Gilbert (1995) call “exploratory simulation”: the objective of the research changes to the observation of and experimentation with possible social worlds. With the possibility of constructing artificial systems, a new methodology of scientific inquiry becomes possible. . . . [T]he exploratory aim synthesizes both the prescriptive and descriptive objectives. (Conte and Gilbert 1995:4) In this project, the descriptive objective of the simulation is to assess the historical impact of the sentencing reforms of the last thirty years on the average dangerousness of prison and other criminal justice populations in California. The prescriptive component consists in experimenting upon the existing system (which has been validated with respect to actual data on criminal justice system populations). The aims of the predictive evaluation component are twofold: the first is to estimate the likely systemic consequences of Three Strikes and other recent sentencing reforms with respect to dangerousness in criminal justice populations. The second objective is to discover ways in which system conditions might be altered so as to maximize the level of dangerousness in the prison population while simultaneously minimizing the level of dangerousness in other criminal justice system populations, such as parole and probation. Despite the naming of this process “predictive evaluation” it bears noting that the analysis is not literally “predictive” in the strictest sense of the word. Byrne (1997) explained the relationship of prediction to simulation thus: “We can’t know what will happen regardless of our acts. We can know what might happen if we act in a certain way” (Byrne 1997:6).
100
Incapacitation and Dangerousness
Some would call this latter aim “social engineering” (e.g., Turner 1998). I resist this label (admittedly) at least in part because of the baggage that accompanies the phrase. However, there are better reasons than ideological squeamishness to reject this label vis-à-vis this project. Rather than imposing a system goal based upon my own moral convictions, I am attempting to find ways to make the system work to better achieve the goals explicitly articulated by its creators and administrators (i.e., the incapacitation of dangerous offenders). For this reason, this project is better understood as an attempt to discover how the state of California might better achieve its own stated goals with respect to criminal sanctioning.
Part III Evaluating the Past, Choosing the Future
This page intentionally left blank.
Chapter 7 Modeling the California Criminal Justice System, Part I Reproducing and Evaluating the Past
This chapter reports the results of the simulation analysis that reproduces the compositional dynamics of the California criminal justice system during the period 1979-1998. The purpose of this chapter is twofold. The first is to demonstrate the validity of the simulation model in terms of its ability to accurately reproduce historical circumstance. The second goal of this chapter is to retrospectively evaluate the California criminal justice system with respect to its success at selectively incapacitating dangerous offenders. Chapter 8 discusses the results of the projection analyses, which are dependent on the plausibility of this baseline model. As chapter 6 explained, dynamic systems simulation modeling is based on the integration of system flows over time. This is accomplished by calculating the values of variables in a complex system of differential and difference equations that are initialized with values based on actual data, and then iterated over time to reproduce an historical sequence of events. System states are represented by differential equations, such as STATE (t) STATE (t1) INFLOWS OUTFLOWS*dt Simply put, the composition of a population state at time (t) is equal to the population at time (t1),1 plus new arrivals, and minus exits. The inflows and outflows are described by rate equations, which represent the proportion of the “sending” state that moves into the “receiving” state at each time step. For example, the equation that models the prison population at time (t) takes the following form:
103
104
Evaluating the Past, Choosing the Future
PRISON (t) PRISON (t1) [(STREET (t) * prison commitment rate from STREET) ( JAIL(t) * prison commitment rate from jail) (PAROLE (t) * return rate)] [ (PRISON(t) * parole rate) (PRISON(t) * unconditional release rate)] *dt In this case, the inflows consist of new commitments to prison resulting from criminal convictions and recommitments resulting from parole violations; the outflows consist of parole releases and unconditional releases. The rate equations are in turn modified by another quantity, which may accelerate or decelerate the rate over time.2 This elaborated form of the equation is PRISON (t) PRISON (t1) [ {(STREET (t) * prison commitment rate from STREET) * street conviction rate modifier} {( JAIL(t) * prison commitment rate from jail) * jail conviction rate modifier} {(PAROLE (t) * return rate) * return rate modifier}] [ {(PRISON(t) * parole rate) * parole rate modifier} {(PRISON(t) * unconditional release rate) * release rate modifier}] *dt Following these principles, a simulation model is estimated by constructing an equation system that codifies the structural relationships depicted in fig-
FIGURE 7.1 Structural Model of the California Criminal Justice System
Modeling the California Criminal Justice System, Part I
105
ure 7.1 for each of the four hundred fifty population groups, and then compiles the results of the subgroups to represent the entire system.3 The modeling of the subpopulations also allows for the estimation of the historical time-shapes of various system components for a variety of populations of interest. In the present analysis, the most pertinent of these is the dangerousness classification scheme; however, this modeling strategy permits the examination of the system histories of a variety of other sorts of populations, such as black men aged 18–24, Hispanic women, and violent offenders—in fact, system histories can be generated for any population characterized by any of the attributes tracked in the simulation model.
THE BASELINE MODEL The following section demonstrates the validity of the simulation model produced by this strategy. Figures 7.2 through 7.6 depict the graphic time series data for each of the system populations modeled, compared with the simulated data for each of those populations.4 One can see that the simulated series mirror the characteristic shapes of the actual data quite closely. In order to provide a target measure for goodness-of-fit, a maximum average
FIGURE 7.2 Arrested Population, 1979–1998
Note: Mean difference between series is 0.2% of target value (standard deviation 6.5%).
FIGURE 7.3 Jail Population, 1979–1998
Note: Mean difference between series is 4% of target value (standard deviation 5.7%).
FIGURE 7.4 Probation Population, 1979–1998
Note: Mean difference between series is 3% of target value (standard deviation 10.7%).
106
FIGURE 7.5 Prison Population, 1979–1998
Note: Mean difference between series is 3% of target value (standard deviation 10.0%).
FIGURE 7.6 Parole Population, 1979–1998
Note: Mean difference between series is 6% of target value (standard deviation 33.7%).
107
108
Evaluating the Past, Choosing the Future
difference of 10 percent of the actual population value was chosen as a critical value. As indicated in the captions to figures 7.2–7.6, each of the simulated series falls within this range.
EVALUATING DANGEROUSNESS IN THE CALIFORNIA CRIMINAL JUSTICE SYSTEM, 1980–19985 It will be recalled that dangerousness is here conceived as a characteristic of populations rather than individuals. Dangerousness in a criminal justice system population is measured using characteristics of the individuals that make up the population. The population dangerousness is calculated as the average dangerousness of individuals in that population. The dangerousness construct presented in chapter 6 (reproduced here in table 7.1) is a simple Burgess-type additive scale that comprises demographic and criminal history information. Included are age, sex, current offense, prior violence history, and number of prior felony convictions. Dangerousness is measured on an eleven-point scale. A one-point difference in dangerousness represents a difference of approximately 9 percent. However, in the same spirit of relativity in which the notion of dangerousness utilized in this research is conceived, it is actually more appropriate to consider differences in population dangerousness (whether comparing changes in one population over time or comparing dangerousness levels between two different populations) with respect to a reference point within the range of values that actually appear, as opposed to the entire range of theoretically possible values. Increases or decreases in dangerousness within populations over time will thus be considered in terms of percent change; comparisons between populations will be considered in percentage terms as well.
Arrested Population One can observe that the drop in the average dangerousness of the arrested population between 1980 and 1998 from a high of 7.34 in 1980 to a low of 6.70 in 1998 represents a decrease of 9 percent in the level of dangerousness in the population of arrestees (see figure 7.7).6 Examination of the data indicates that this slight but steady decline can be explained by a proportional increase over time in arrests of older offenders, with corresponding decreases in the proportion of young (18–24) offenders (see figure 7.8). Additionally, the proportion of arrests for drug offenses has increased in this period. The proportion of arrests for violent offenses has also increased, while arrests for property offenses as a proportion of all arrests has steadily declined over this period (see figure 7.9).
Modeling the California Criminal Justice System, Part I
109
TABLE 7.1 Coding Scheme for Dangerousness Measure Characteristic
Score
Age 18–24 25–34 35–44 45–54 55 and older
4 3 2 1 0
Male Female
1 0
2 1 none
2 1 0
1 or more none
1 0
violent property drug
3 2 1
Gender
Prior felony convictions
Prior violent convictions
Current conviction offense
Minimum score: 1 (55 year-old woman, no priors, drug offense) Maximum score: 11 (18–24 year-old male, 2 or more priors, violent prior[s], violent offense)
As figures 7.10 and 7.11 demonstrate, the relative proportions of prior conviction and violence histories have not changed significantly over this period;7 it therefore appears that the primary influences on the declining dangerousness of the arrested population are the changes in the age and offense distribution in the population of arrestees.
Jail Population Figure 7.12 shows that after an initial slight increase in dangerousness from 7.07 in 1979 to a high of 7.26 in 1983, there has been a 6 percent decrease in the average dangerousness of the jail population from 1983 through 1998. This decline may be explained by much the same pattern that appears in the arrests; while the proportion of jail population held for violent offenses has
FIGURE 7.7 Dangerousness of Arrested Population, 1980–1998
FIGURE 7.8 Age Distribution of Arrested Population, 1980–1998
110
FIGURE 7.9 Offense Distribution of Arrested Population, 1980–1998
FIGURE 7.10 Distribution of Prior Felony Convictions, Arrested Population, 1980–1998
111
FIGURE 7.11 Distribution of Prior Violence History, Arrested Population, 1980–1998
FIGURE 7.12 Dangerousness of Jail Population, 1980–1998
Modeling the California Criminal Justice System, Part I
113
remained relatively constant, property offenders are increasingly being replaced by drug offenders (see figure 7.14). Similarly, younger offenders (those aged 18–24) comprise an increasingly smaller proportion of jail populations, while there is a corresponding increase in middle-aged and older offenders (see figure 7.13). While the proportion of offenders in jail with no prior convictions remains more or less constant over the period, the proportion of offenders with two or priors has increased, with a corresponding decrease in the proportion of offenders with a single prior conviction (see figure 7.15). At first glance, this would seem to be consistent with Turner’s (1999) finding that Three Strikes offenders were taking the place of sentenced offenders in Los Angeles jails, due to increased rates of pretrial detention in response to the greater flight risk posed by these defendants. However, the increase in offenders with more lengthy criminal histories precedes the Three Strikes law by a number of years. It is therefore more likely that this shift in the composition of the jail population is attributable to other causes, such as the increased use of mandatory-minimum drug statutes throughout the 1980s. Following the same logic as Turner’s analysis of Three Strikes, it is conceivable that offenders facing lengthy mandatory minimums would represent an increased flight risk, and thus might be more likely to be remanded to detention prior to adjudication.8 As figure 7.16 shows, the distribution of offenders according to prior violence history has remained static over the period 1980–1998.
Probation Population Figure 7.17 indicates that the probation population has become increasingly dangerous over time. As figure 7.17 shows, the dangerousness level of the probation population peaked in the mid- to late- 1980s, and declined thereafter, but remained at a higher level than that of the early part of the 1980s. In quantitative terms, the dangerousness of the probation population increased 31% between 1980 and 1987, and declined 9% between 1987 and 1998. The net change in the dangerousness of the probation population from a value of 7.6 in 1980 to a value of 9.01 in 1998 is an increase of 19%. Possible explanations for this increase in the dangerousness of the probation population may lie in the changes in sentencing combined with capacity limitations in jails that occurred throughout the 1980s. As penalties for drug offenses became increasingly harsh, a trade-off relationship may have developed whereby less dangerous drug offenders occupied available jail space (either as pretrial detainees, due to the perception of flight risk, or as sentenced offenders), resulting in more dangerous offenders who might have otherwise received a jail sentence being placed on probation.9
FIGURE 7.13 Age Distribution of Jail Population, 1980–1998
FIGURE 7.14 Offense Distribution of Jail Population, 1980–1998
114
FIGURE 7.15 Distribution of Prior Felony Convictions, Jail Population, 1980–1998
FIGURE 7.16 Distribution of Prior Violence History, Jail Population, 1980–1998
115
116
Evaluating the Past, Choosing the Future FIGURE 7.17 Dangerousness of Probation Population, 1980–1998
Figure 7.19 seems to support this notion. The distribution of conviction offenses in the probation population has changed dramatically over the last two decades. While drug offenders comprise the greatest fraction of the probation population over the entire period, the relative positions of property and violent offenders have flipped. Since 1980, the proportion of violent offenders in the probation population has increased 36%, while the proportion of drug offenders has increased only 15%, and property offenders as a share of the probation population have declined 38%. Figure 7.18 indicates that changes in the age distribution of the probation population have been minimal, and are not in a direction that would contribute significantly to the increased dangerousness of the probation population. As Figure 7.21 indicates, the proportion of probationers with histories of violence has increased 20% from 1980 to 1998; this is consistent with the increase in probationers with violent conviction offenses. Figure 7.20 shows an increase in the proportion of probationers with multiple priors. This is also consistent with the displacement scenario outlined above.
Prison Population During the period 1980–1998, the average level of dangerousness in the prison population has shown a steady 8% decline, from a high of 7.58 in
FIGURE 7.18 Age Distribution of Probation Population, 1980–1998
FIGURE 7.19 Offense Distribution of Probation Population, 1980–1998
FIGURE 7.20 Distribution of Prior Felony Convictions, Probation Population, 1980–1998
FIGURE 7.21 Distribution of Prior Violence History, Probation Population, 1980–1998
Modeling the California Criminal Justice System, Part I
119
1980 to a low of 7.01 in 1998 (figure 7.22). This decrease can be explained primarily by the aging of the prison population over this period, and by the increase in drug offenders relative to other offenders (see figures 7.23 and 7.24, respectively). The changes in the age distribution of the incarcerated population (demonstrated in figure 7.23) are striking. While the youngest offenders (those aged 18–24) comprised 31% of the prison population in 1980, only 22% of the prison population was under 25 in 1998. Similarly, the percentage of offenders aged 25–34 declined from 47% to 37% during this period. The proportion of older offenders increased dramatically during this period; the proportion of offenders aged 45–54 doubled, to become 10% of the prison population in 1998. While less than 25% of the prison population was over 35 years old in 1980, nearly 50% of prisoners were older than 35 in 1998. Even more remarkable are the changes in the distribution of conviction offenses in the prison population during this period. While in other criminal justice system populations we observe a pattern of change that appears to be a trading-off between drug and property offenses, with the proportion of violent offenses remaining relatively constant, the change in the distribution of offenses in the prison population from 1980 to 1998 follows rather a different
FIGURE 7.22 Dangerousness of Prison Population, 1980–1998
FIGURE 7.23 Age Distribution of Prison Population, 1980–1998
FIGURE 7.24 Offense Distribution of Prison Population, 1980–1998
Modeling the California Criminal Justice System, Part I
121
pattern. While the proportion of property offenses has remained relatively constant—around 25% of the prison population—the proportion of violent offenders has declined sharply. As figure 7.24 demonstrates, while violent offenders comprised nearly two-thirds of the prison population in 1980, by 1998 less than half of the prison population consists of offenders serving a sentence for a violent conviction offense. In 1980, barely 10% of California prisoners were incarcerated on a drug offense; in 1998, this figure is over 30%. The proportion of prisoners with lengthy criminal histories (two or more prior convictions) has sharply increased, to more than double 1980 levels in 1998 (see figure 7.25); this increase in more dangerous offenders mitigates the impact of the demographic changes on the declining dangerousness of the prison population. This increase is clearly not attributable to the 1994 Three Strikes law, as the trend in the series predates the law’s passage. While the relative proportion of violent and nonviolent offenders in the prison population has demonstrated some minor fluctuation over time, the distribution of offenders by violence history has remained essentially unchanged over the last two decades (figure 7.26). Much has been made of increased rates of the incarceration of women relative to men in California in recent years (e.g., Bloom, Chesney-Lind,
FIGURE 7.25 Distribution of Prior Felony Convictions, Prison Population, 1980–1998
FIGURE 7.26 Distribution of Prior Violence History, Prison Population, 1980–1998
FIGURE 7.27 Rate of Growth in Male and Female Prison Population, 1980–1998
Modeling the California Criminal Justice System, Part I
123
and Owen 1994). The historical data indicate that this disparity has been considerably overstated. Because the relative sizes of the male and female prison populations are so disparate, figure 7.27 displays the annual percentage growth in the size of each of these populations for relatively greater ease of visual interpretation. While the rate of growth in California’s female prison population outpaced that of males in the early 1980s, the rates of growth converge thereafter, averaging around a 15% rate of growth for females and 12% for males over the entire period.10 In light of this, it seems unlikely that the increased incarceration of women has contributed significantly to the declining dangerousness of the prison population. Two things should be noted about the dangerousness analysis of the prison population. While an 8% decrease in dangerousness is not a dramatic decline, it should be evaluated with respect to the sweeping array of criminal sentencing reform largely aimed at selective incapacitation and the subsequent explosive growth of the prison population that has also taken place during this period. From the standpoint of selective success, these polices appear to have failed miserably at their task. If selective incapacitation policies have been successful at targeting and incapacitating dangerous offenders, we should observe a net increase in the average level of dangerousness in the prison population, rather than a decrease. The second point that should be noted is that the most sweeping of these reforms—namely, Three Strikes—has not been in effect long enough to show the consequences to prison populations. The projection analyses that follow in chapter 8 will provide valuable insight into the likely future consequences of this and other recent reforms.
Parole Population As figure 7.28 shows, there has been essentially no net change in the dangerousness of the parole population from 1980 through 1998; the series begins with a value of 6.05 in 1980 and ends with a value of 6.07 in 1998. Over this period, however, the dangerousness of the parole population increased 9 percent in the early 1980s and then declined, arriving back at 1980 levels by the late 1990s. Not surprisingly, changes in the composition of the parole population reflect changes in the prison population. Figure 7.29 demonstrates that the aging of the parole population mirrors that of the prison population. As figure 7.30 indicates, the proportion of drug offenders has nearly doubled, with decreases of about equal magnitude in the proportions of violent and property offenders. Figure 7.31, which depicts the distribution of prior convictions in
FIGURE 7.28 Dangerousness of Parole Population, 1980–1998
FIGURE 7.29 Age Distribution of Parole Population, 1980–1998
Modeling the California Criminal Justice System, Part I
125
FIGURE 7.30 Offense Distribution of Parole Population, 1980–1998
the parole population, shows a more gradual increase in the proportion of offenders with two or more convictions, mirroring the composition of the prison population. The proportion of parolees with a prior violent conviction has slightly increased over the period (figure 7.32), but it seems unlikely that this has had any significant impact on the overall dangerousness of the parole population. The points made about the prison population bear repeating here. It is somewhat disturbing to observe that two decades of reform aimed at incapacitating dangerous criminals, primarily by targeting certain types of offenders for more lengthy incarceration, has had virtually no effect on the dangerousness level of the parole population. If increased and extended incarceration were successful at incapacitating dangerous offenders, we might expect to see a decreased level of dangerousness in the parole population (preferably in tandem with an increased level of dangerousness in the prison population). Finally, the observation offered above regarding the recency of reforms is also applicable here. The effects of Three Strikes and other reforms have not had enough time to manifest themselves, but will be more easily identified in the projection analyses.
FIGURE 7.31 Distribution of Prior Felony Convictions, Parole Population, 1980–1998
FIGURE 7.32 Distribution of Prior Violence History, Parole Population, 1980–1998
Modeling the California Criminal Justice System, Part I
OF
127
COMPARING DANGEROUSNESS CRIMINAL JUSTICE POPULATIONS
A criminal justice system functioning according to the principles of selective incapacitation should exhibit certain properties. From a systemic perspective, dangerousness should be maximized in incarcerated populations, and minimized in nonincarcerated populations if policies intending to achieve the selective incapacitation of dangerous offenders are successful. Examining the results presented in figure 7.33, it is apparent that this is has not been the case in California in the last two decades. While the prison population is nearly the most dangerous of the criminal justice system populations in 1980 (the probation population has a dangerousness value that is minutely larger than that of the prison population), after 1980, this is no longer the case. The probation population is the most dangerous criminal justice system population over the entire period. Indeed, after 1980, the probation population is, on average, 32% more dangerous than the prison population, and 35% more dangerous than the jail population. The arrested population is less dangerous throughout the series than all other criminal justice system populations, except parole. This is as it should be, as the process of arrest would seem to exhibit the least selectivity—that is, arrest
FIGURE 7.33 Relative Dangerousness of Criminal Justice Populations, 1980–1998
128
Evaluating the Past, Choosing the Future
casts the widest net, and then the system should, theoretically, filter and select more serious offenders for sanctions. However, comparing the dangerousness levels in the arrested population and the jail population indicates that the jail population is, on average, only marginally more dangerous than the arrested population as a whole (in 1980 the jail population is actually 4% less dangerous than the arrested population, and thereafter exhibits a dangerousness level that is approximately 2% greater than that of the arrested population). Even more striking, the prison population is also only marginally more dangerous than the arrested population during this period; on average, the prison population is only 4% more dangerous than the arrested population. The prison population is at all points more dangerous than the parole population. This is consistent with the successful implementation of selective incapacitation policies, but the difference in dangerousness between the two system populations is not large. While the prison population was 25% more dangerous than the parole population in 1980, the gap between the two narrows thereafter, with the dangerousness of the prison population averaging only 13% more than that of the parole population.
CONCLUSION From the standpoint of selective incapacitation, the results of the retrospective evaluation of the selective success of the California criminal justice system are uniformly disappointing. The dangerousness of incarcerated populations has decreased rather than increased in the last two decades, while the dangerousness of noncustodial populations has either remained relatively constant (parole) or increased dramatically (probation). The analyses of the individual components of the dangerousness measure indicate that the factors exerting the most influence on the dangerousness of criminal justice system populations are age and current offense. The principal trends are the aging of criminal justice populations and the increased influx of drug offenders into the system. While the decreases in dangerousness of the jail and prison populations are not large, they assume a greater significance in light of the two decades of sentencing policy reform primarily directed at increasing the public safety through the incapacitation of dangerous offenders. While the emphasis in this work has not been on the numerical growth of California’s criminal justice system, it behooves us to take this into account in evaluating the success of criminal sentencing policies in terms of protecting the public from dangerous offenders. Put another way, it should be noted that as the results of the preceding analysis indicate that policies intended to make the people of California safer via the incapacitation of dangerous offenders have not been terribly successful, these not-so-successful policies have also had the
Modeling the California Criminal Justice System, Part I
129
added effect of expanding the state’s entire criminal justice infrastructure. The magnitude of this expansion varies according to the area of the criminal justice system. The prison population has undergone a sixfold expansion, jail and probation populations have tripled, and the parole population has increased eightfold since 1980. Accordingly, annual criminal justice expenditures in the late 1990s have increased more than five times, climbing from $3 billion in 1979 to a staggering $18.7 billion in fiscal year 1998–1999 (California Department of Justice 1981, 2002). At the “back-end” of the sanctioning system, it appears that parole is functioning reasonably well in terms of selective incapacitation, but it is apparent that the system as a whole is not effectively keeping pace with the massive influx of drug offenders into the system since the mid-1980s. Most disturbing is the pattern of change in the distribution of conviction offenses in the prison population. While in the other criminal justice system populations there appears to be a process of substitution of drug offenders for property offenders, in the prison population, the proportional representation of property offenders has remained relatively constant over the period 1980–1998. What has instead happened in the prison population is the displacement of violent offenders by drug offenders. These displaced violent offenders increasingly appear in noncustodial criminal justice system populations, notably probation. This analysis highlights the importance of analyzing criminal justice reforms from a systemic perspective. While the vast majority of criminal justice reform in the past two decades has been directed at altering the composition of the prison population via the selective incapacitation of dangerous offenders, the analyses reported in this chapter demonstrate that these reforms have farreaching effects on other areas of the criminal justice system, such as the jail and probation systems. How might California’s criminal justice system do better at protecting its citizens through selective incapacitation? The analyses reported in this chapter indicate that criminal sentencing policy as it currently operates is doing rather poorly at fulfilling this promise. Chapter 8 reports the results of projection analyses designed to determine whether these policies can be altered in a way that more fully achieves the goals of selective incapacitation.
This page intentionally left blank.
Chapter 8 Modeling the California Criminal Justice System, Part II Predictive Evaluation
Chapter 7 reported the results of the retrospective simulation analyses evaluating the efficacy of criminal sentencing reform in California with respect to the objective of incapacitating dangerous offenders. As was noted in those analyses, one of the most far-reaching reforms, namely, the state’s Three Strikes law, has not yet begun to show its effects on the criminal justice system. For this reason, this study utilizes an approach I call predictive evaluation. Predictive evaluation entails simulation of the effects of the implementation of various policy schemes, and the evaluation of these schemes with respect to their success at selectively incapacitating dangerous offenders. The predictive evaluation approach is useful for evaluating not only Three Strikes, but all kinds of recent or proposed sentencing innovations, particularly reforms that are specifically designed to lengthen the term of incarceration for certain types of offenders. Such reforms take several years for their effects to materialize, in that the majority of offenders affected by such a law would in all likelihood have been incarcerated under the sentencing structure existing prior to the implementation of the law (see Kovandzic 2001:181). In a retrospective intervention analysis, the effects of a reform may be estimated by lagging the outcome variable by some period of time believed to represent the time to implementation of the law. In the case of Three Strikes, data are not yet available for a sufficient length of time postintervention to conduct reliable statistical analyses. Simulation modeling also allows us to evaluate the likely consequences of other “possible futures.” In this way, we can compare alternative scenarios in order to determine which policies are best suited for achieving the objective of incapacitating dangerous offenders. The following analyses project the simulation model developed in the chapter 7 forward in time to the year 2015 under a 131
132
Evaluating the Past, Choosing the Future
variety of conditions, in order to gain insight into the possible consequences of different sentencing schemes. In the interest of brevity, and because the prison is the most central component of the criminal justice system for the incapacitation of dangerous offenders, this chapter will focus on the demographic consequences to the prison population of the various policy options. Four possible sentencing structures are examined. These include three variants on California’s Three Strikes law, and another sentencing innovation that has received some attention in policy circles in recent years, geriatric release. This entails the release of elderly offenders prior to the expiration of their minimum terms.
SCENARIO 1: THREE STRIKES, F ULL IMPLEMENTATION The first scenario investigated simulates the likely future consequences of the continued full implementation of California’s Three Strikes law on the prison population. The Three Strikes law as written provides for the mandatory doubling of the presumptive sentences for offenders with one strike and a second strikeable felony, and a mandatory tripling of the presumptive term (or twenty-five years to life, whichever is longer) for offenders with two strikes, convicted of any felony. The implementation of the law was simulated by reducing parole release rates for all offenders with a prior felony conviction, and to simulate the effect of the exceptionally long sentences imposed on some third-strike offenders, also reducing the rate of direct release from prison for all offenders with two or more prior convictions.1 As figure 8.1 demonstrates, full implementation of the Three Strikes law will not change the overall dangerousness level of the California prison population, although the preexisting decline does appear to level off somewhat; the dangerousness level in the year 2015 is approximately 1% lower than that in 1998. The components of the dangerousness construct behave in predictable ways given the trends demonstrated in chapter 7 and the features of the Three Strikes habitual offender law. Figure 8.2 indicates that under these conditions, the proportion of older offenders will increase substantially, a foreseeable consequence of a law designed to incarcerate certain offenders for lengthy terms. Under the conditions of full implementation of Three Strikes as written, the proportion of prisoners aged 55 and older can be expected to increase 33% by 2015, and the proportion of those aged 45–54 will increase 50% compared to 1998. Consistent with early reports that the California law disproportionately affects nonviolent drug offenders (Austin et al. 1999; Legislative Analyst’s Office 1999), it appears that the proportion of offenders incarcerated for drug offenses will increase 6%, while the proportional representation of property offenders will decrease 12.5%. The representation of violent offenders is expected to remain relatively constant over the period (figure 8.3). Not
FIGURE 8.1 Dangerousness of Prison Population, 1980–2015: Three Strikes, Full Implementation
FIGURE 8.2 Age Distribution of Prison Population, 1980–2015: Three Strikes, Full Implementation
FIGURE 8.3 Offense Distribution of Prison Population, 1980–2015: Three Strikes, Full Implementation
FIGURE 8.4 Distribution of Prior Felony Convictions, Prison Population, 1980–2015: Three Strikes, Full Implementation
Modeling the California Criminal Justice System, Part II
135
surprisingly, Figure 8.4 shows that following the implementation of the Three Strikes law, the proportion of offenders with two or more prior convictions will increase by 30%. It is important to note, however, that this increase appears to capitalize on a preexisting trend—the proportional representation of offenders with lengthy criminal histories has steadily increased since the early 1980s (see chapter 7). It would be disingenuous to attribute this entire increase to the Three Strikes law. As figure 8.5 demonstrates, the relative proportions of offenders with and without histories of violence remains essentially unchanged, exhibiting a slight increase in the proportion of violent offenders in the prison population.2 In sum, the Three Strikes law appears to cause some changes in the composition of the prison population favorable to increasing the dangerousness of the population (such as increases in the proportion of offenders with violent histories, violent conviction offenses, and lengthy criminal histories). However, these benefits appear to be all but completely offset by the aging of the prison population. Put another way, these analyses seem to indicate that the Three Strikes law doesn’t appreciably increase the selective success of the prison—it is largely the same offender pool that is present in recent years—Three Strikes just keeps them there longer.
FIGURE 8.5 Distribution of Prior Violence History, Prison Population, 1980–2015: Three Strikes, Full Implementation
136
Evaluating the Past, Choosing the Future
It is often useful to note other demographic consequences to the prison population, particularly the impact (if any) of particular policy choices on the racial distribution of the prison population. Observers have noted that the earliest data available on the implementation of the Three Strikes law appear to demonstrate a disproportionately harsh effect on African-Americans (Schiraldi with Godfrey 1994; Davis, Estes, and Schiraldi 1996). Figure 8.6 offers some support for this contention, although the effects of the law as modeled are not as dramatic as some have predicted. The simulation results indicate that the implementation of California’s Three Strikes law has the effect of leveling or reversing a downward trend in the proportional representation of AfricanAmericans in the prison population, resulting in a 3% increase (from 1998 levels) in the relative size of the African-American prison population. The proportion of the prison population that is Hispanic is expected to grow 6% by 2015, consistent with preexisting trends, while the proportional representation of white offenders will decrease 5%. What is notable is not the relatively small proportional shifts over the simulated period, but rather the possible long-term consequences of the reversal of the downward trend in the representation of African-American prisoners.
FIGURE 8.6 Racial Distribution of Prison Population, 1980–2015: Three Strikes, Full Implementation
Modeling the California Criminal Justice System, Part II
137
SCENARIO 2: THREE STRIKES, VIOLENT OFFENSE ONLY In addition to simulating the likely consequences of the implementation of California’s Three Strikes law as written, two alternate Three Strikes scenarios were also simulated. The first of these involves limiting the eligibility for the Three Strikes provisions to only those offenders with a violent conviction offense (and the requisite number of prior felony convictions).3 The results of this simulation are depicted in figures 8.7–8.12. As figure 8.7 shows, altering the law in this way similarly halts the decline in dangerousness. This modification almost completely stops the decline, with the 2015 dangerousness level exhibiting a decrease of less than 1% from 1998 levels. The age distribution pattern depicted in figure 8.8 is nearly identical to that exhibited under the full-implementation scenario; the proportion of offenders 55 and older increases 33% from 1998 to 2015, while the increase in the proportion of offenders aged 45–54 is slightly less than the full implementation scenario (45%). Figure 8.10 depicts a much less dramatic increase of 7% in the proportion of offenders with lengthy criminal histories (two or more prior felony convictions) than in the full-implementation scenario, a predictable consequence
FIGURE 8.7 Dangerousness of Prison Population, 1980–2015: Three Strikes, Violent Only
FIGURE 8.8 Age Distribution of Prison Population, 1980–2015: Three Strikes, Violent Only
FIGURE 8.9 Offense Distribution of Prison Population, 1980–2015: Three Strikes, Violent Only
138
FIGURE 8.10 Distribution of Prior Felony Convictions, Prison Population, 1980–2015: Three Strikes, Violent Only
FIGURE 8.11 Distribution of Prior Violence History, Prison Population, 1980–2015: Three Strikes, Violent Only
139
140
Evaluating the Past, Choosing the Future
given the narrowing of eligibility for Three Strikes. Figure 8.11 indicates that under this scenario, the proportion of offenders with violent histories remains essentially the same from 1998 to 2015. Not surprisingly, the primary difference between this and the implementation scenario for all offenders with prior convictions is the increase in the proportion of offenders serving sentences for violent crimes; in 2015, under a Three Strikes policy that targeted only offenders with a violent conviction offense, over 50% of the California prison population would consist of violent offenders—bringing the proportion nearly back to 1980 levels (figure 8.9). The consequences of restricting Three Strikes eligibility to only those offenders with a violent conviction offense has an impact on the racial composition of the California prison population that is similar to the full implementation of Three Strikes. As figure 8.12 demonstrates, while the proportion of white prisoners exhibits a small decrease, it appears that the representation of African-Americans and Hispanics would grow, by 3% and 6%, respectively. It should be noted that these sorts of findings do not necessarily indicate that the sentencing structure is discriminatory. Rather, the analyses of the racial composition of the prison population under the various scenarios is offered merely to show that policy choices may have unforeseen systemic effects that are not necessarily neutral with respect to their impact on different communities.
FIGURE 8.12 Racial Distribution of Prison Population, 1980–2015: Three Strikes, Violent Only
Modeling the California Criminal Justice System, Part II
141
SCENARIO 3: THREE STRIKES, VIOLENT HISTORY ONLY The third variant of Three Strikes explored in this analysis represents a middle ground—in between full implementation and stringent restriction of eligibility to only those with a violent strike. This third scenario estimates the effects of applying Three Strikes eligibility to all offenders with a prior conviction and a violent strike, and to any offender with two or more convictions and a history of violence, regardless of current conviction offense.4 The results of these analyses are depicted in figures 8.13–8.18. Figure 8.13 shows that this third Three Strikes scenario would also keep the dangerousness level of the prison population at 1998 levels. The consequences to the age distribution of the prison population are virtually identical to the other Three Strikes scenarios—proportional representation of the youngest offenders will continue to decrease between 1998 and 2015, and the proportion of older offenders 45–54, and 55 and up will increase by 40% and 33%, respectively (figure 8.14). Figure 8.15 shows that this wider strike eligibility zone increases the proportion of violent offenders 11% over the period 1998–2015, somewhat less than the 16% increase observed in the violent-only scenario just discussed. The proportions of drug and property offenders show slight declines. The narrower strike
FIGURE 8.13 Dangerousness of Prison Population, 1980–2015: Three Strikes, Violent History
FIGURE 8.14 Age Distribution of Prison Population, 1980–2015: Three Strikes, Violent History
FIGURE 8.15 Offense Distribution of Prison Population, 1980–2015: Three Strikes, Violent History
142
FIGURE 8.16 Distribution of Prior Felony Convictions, Prison Population, 1980–2015: Three Strikes, Violent History
FIGURE 8.17 Distribution of Prior Violence History, Prison Population, 1980–2015: Three Strikes, Violent History
143
144
Evaluating the Past, Choosing the Future FIGURE 8.18 Racial Distribution of Prison Population, 1980–2015: Three Strikes, Violent History
zone results in a 19% increase in offenders with two or more prior convictions by 2015 (seen in figure 8.16). Predictably, this is a much greater increase than the 7% increase in the most restrictive scenario, and less than the 30% increase seen in the full implementation scenario. The most dramatic effect with respect to dangerousness is the 13% increase in the proportion of offenders with violent histories. As figure 8.17 indicates, the implementation of this variant of Three Strikes would result in nearly half of incarcerated offenders having histories of violence. This stands in stark contrast to the Three Strikes law currently operating, which incarcerates large numbers of offenders with no history of violence whatsoever. Figure 8.18 indicates that this scenario would engender consequences similar to the other Three Strikes variants with respect to the racial composition of the prison population, namely, an increase in African-American and Hispanic inmates, and a corresponding decease in white inmates.
SCENARIO 4: GERIATRIC RELEASE Another policy option that might improve the selective success of incarceration in California is geriatric release. This entails the release of older offenders prior
Modeling the California Criminal Justice System, Part II
145
to the end of their minimum terms, based on the presumption that such offenders have “aged out” of criminal behavior and therefore no longer represent a significant threat to society.5 It should also be noted that it is potentially advantageous from a financial standpoint for the department of corrections to release these offenders, as it can cost up to three times more yearly to maintain an elderly offender than a younger offender in prison. This is primarily due to the increased health-care needs of these offenders (Zimbardo 1994:3).6 Figures 8.19 through 8.24 depict the results of a simulation that superimposes a program of geriatric release on to the “full implementation” Three Strikes scenario. For the purposes of the simulation, “geriatric” was defined as inmates 55 or older. While 55 is much younger than what we normally consider elderly, from the standpoint of aging out of criminal behavior, criminologists are generally in agreement that 55 is well past the most active offending years (Petersilia 1980; Farrington 1986; Auerhahn 2002).7 The geriatric release scenario involves increasing the release rates of offenders aged 55 and older until such offenders comprise a negligible fraction of the prison population. Figure 8.19 shows that this scenario (geriatric release with full Three Strikes implementation) would have virtually the same consequences to the dangerousness of the prison population as the full implementation of Three
FIGURE 8.19 Dangerousness of Prison Population, 1980–2015: Geriatric Release Scenario
FIGURE 8.20 Age Distribution of Prison Population, 1980–2015: Geriatric Release Scenario
FIGURE 8.21 Offense Distribution of Prison Population, 1980–2015: Geriatric Release Scenario
146
FIGURE 8.22 Distribution of Prior Felony Convictions, Prison Population, 1980–2015: Geriatric Release Scenario
FIGURE 8.23 Distribution of Prior Violence History, Prison Population, 1980–2015: Geriatric Release Scenario
147
148
Evaluating the Past, Choosing the Future FIGURE 8.24 Racial Distribution of Prison Population, 1980–2015: Geriatric Release Scenario
Strikes—a dramatic slowing of the decrease of dangerousness, but no real gains in terms of increasing dangerousness relative to 1998 levels. A closer inspection of figure 8.19 shows that the dangerousness construct declines somewhat after 1998, but rebounds and levels off by 2001, as the elderly prisoners exit the prison population. Figure 8.20 indicates that this will not dramatically change the age structure of the prison population, save for the 90% decline in the representation of prisoners 55 and older. The proportional size of this population declines from 3% in 1998 to .3% in 2015. Apart from this difference, figures 8.20 through 8.24 indicate that implementing a geriatric release program alongside the full implementation of Three Strikes will manifest similar consequences to the first scenario discussed. The configurations of criminal history, conviction offense, violent history, and racial composition are otherwise identical to those observed under the full implementation of Three Strikes.
COMPARING POSSIBLE F UTURES Table 8.1 offers a side-by-side comparison of the big picture with respect to selective incapacitation under the four policy scenarios simulated. It appears that
Modeling the California Criminal Justice System, Part II
149
the variant of Three Strikes that modifies the strike zone to include all violent offenders with at least one prior conviction, and all offenders with at least one prior violent conviction performs the best in terms of maximizing the level of dangerousness in the prison population. It is also apparent that either of the two scenarios that model modifications to the strike zone would perform marginally better at incapacitating dangerous offenders than the Three Strikes law as written. Geriatric release combined with full implementation of Three Strikes has virtually no net effect on dangerousness, and in fact, this scenario slightly decreases dangerousness (as does full implementation Three Strikes).8 Table 8.1 shows the consequences to prison population growth posed by each of the four possible future scenarios. Here again, it is apparent that any of the alternatives demonstrated would perform better than the Three Strikes law as currently written, assuming that curtailing the explosive growth in the prison population is seen as desirable. In 1998, the California prison population stood at approximately 155,998 inmates. Under the conditions of full implementation of the Three Strikes law, this rate of growth will accelerate, increasing the prison population nearly 60% by 2015. Under the most restrictive scenario, that which limits Second- and Third-Strike eligibility to only those with a violent conviction offense, the prison population would grow to approximately 216,426 inmates by 2015, an increase of approximately 39% from 1998. The middleground scenario with respect to eligibility is also predictably the middle ground with respect to prison population growth—under the violent-history scenario, the prison population in 2015 is estimated at 226,872—an increase of 45% over the 1998 population.
TABLE 8.1 Comparing Future Scenarios Dangerousness
1: 2: 3: 4:
Full implementation Violent only Violent history Geriatric release
Prison Population
1998*
2015
change
1998**
2015
change
6.97 6.97 6.97 6.97
6.90 6.94 7.01 6.87
–1.0% –0.4% 0.6% –1.4%
155,998 155,998 155,998 155,998
247,592 216,423 226,872 239,128
58.7% 38.7% 45.4% 53.2%
*The dangerousness value varies slightly for each simulation run. This is because each new iteration generates the system spontaneously, resulting in minor variations. The value given here is the mean of the four values. The standard deviation is .07. **The population value is also slightly different for each simulation run. The value given here is the mean of the four values. The standard deviation is 150.6 prisoners.
FIGURE 8.25 Comparison of Prison Population Growth, 1998–2015
Modeling the California Criminal Justice System, Part II
151
Figure 8.25 depicts these population growth curves. This visual representation provides another window into the long-term potential consequences of these policies. The full implementation of Three Strikes, with or without geriatric release (scenarios 1 and 4) appear to accelerate the rate of prison population growth. The scenario that simulates the narrowest Second- and Third-Strike zone (number 2), for violent instant offenses only exhibits a slight shallowing of the curve, or a deceleration of the rate of growth. The violent-history eligible scenario (number 3)—more restrictive than the law as written but encompassing a broader definition of “violent offender”—appears to continue the existing rate of growth, rather than accelerating or decelerating it.
CONCLUSION It is apparent that any variant of a Three Strikes law will have two consequences, consistent with preexisting trends: increasing the proportion of offenders with substantial criminal histories in prison, and increasing the proportion of older offenders in the prison population. This is a necessary byproduct of a policy that subjects some portion of offenders to lengthy terms of incarceration. However, the benefits that accrue with respect to the dangerousness of the prison population by raising the proportion of offenders with multiple prior convictions are neutralized by the aging of the prison population. If Californians decide to accept this consequence, viewing a sentencing structure that provides for lengthy sentences for offenders with substantial criminal histories as beneficial,9 then other options (such as narrowing the strike zone) must be explored if dangerousness is to be maximized in the prison population. It is clear from the foregoing analyses that the path that California sentencing policy is currently following will have disappointing consequences from the standpoint of selective incapacitation. The Three Strikes law currently in effect will result in substantial growth in the prison population, for relatively little benefit vis-à-vis the overall dangerousness of the prison population. Other scenarios explored clearly offer (marginally) greater benefit, whether those scenarios involve limiting strike eligibility or taking steps to curtail the aging of the prison population. However, we would be missing the point were we to only consider these findings in and of themselves. These analyses are valuable for what they do not demonstrate. As wide and sweeping as the potential consequences of California’s Three Strikes appear in the abstract, they pale in comparison to the cumulative effects of the changes made to California sentencing policy during the period 1980–1998. The damage, it seems, is already done. With respect to the selective incapacitation of dangerous offenders, the great expenditures associated with further
152
Evaluating the Past, Choosing the Future
increases in the prison population via the Three Strikes law will not buy much. Realistically, the most that we can expect from these changes is to keep things from getting appreciably worse. Chapter 9 assesses the results of these analyses in terms of how the knowledge gained from this exercise can and should be used by policymakers interested in ensuring the public safety through the incapacitation of dangerous offenders.
Chapter 9 Conclusion Choosing California’s Future
This work has attempted to explicate the linkages between the goals and consequences of public policy. The analyses have focused on criminal sentencing policy reform in late-twentieth-century California. Using the extremely powerful tool of simulation analysis to model California’s criminal justice system in past and future, I have endeavored to highlight the pernicious unintended consequences that result from the actual operation of policies which, in their inception, are ostensibly well-intentioned. As the analyses reported in the chapters 7–8 have shown, two decades of sentencing policy reforms conceived and implemented with the goal of making California’s citizens safer have in fact resulted in a configuration of offenders at various levels of criminal justice system supervision that may indeed place the public at ever-greater levels of risk. The projection analyses offered in chapter 8 demonstrate that Californians find themselves at a crossroads: faced with the choice of pushing onward along the path already chosen, or of making bold steps toward innovation in order to achieve the originally intended goals. This is the purest virtue of simulation analysis; it allows us to formulate a number of likely possible futures and to envision the consequences of those possibilities. We need not simply wait for the future to overtake us; we have choices. One contention of this work is that the most prominent promise of criminal sentencing policy reform in California in recent years has been to protect the public from dangerous offenders. As the retrospective analyses have shown, California has failed miserably in fulfilling this promise. The explanation for this is simple: broadly written get-tough sentencing policies have far-reaching effects on the criminal justice system, which ultimately result in unintended consequences. Systems theorist Jay W. Forrester (1969b) once observed, “Intuition is unreliable. 153
154
Evaluating the Past, Choosing the Future
It is worse than random because it is wrong more often than not when faced with the dynamics of complex systems” (Forrester 1969b:24). An understanding of the effects and dynamics of the operation of complex systems is essential in order to anticipate the potential consequences of reforms, especially dramatic, sweeping reforms such as California’s Three Strikes law. Faced with the findings about the likely consequences of this overhelmingly popular reform as reported in chapter 8, I am put in mind of the words of Bertrand Russell (1950): “most of the greatest evils that man has inflicted upon man have come about through people feeling quite certain about something which, in fact, was false” (Russell 1950:162). While the primary focus of this work has been that of evaluation of policy from the standpoint of efficacy with respect to policy goals, it should not go without saying that sentencing structures that have the effect of incarcerating large numbers of nonviolent, elderly, and disproportionately minority offenders exact a substantial human toll, in addition to the financial and operational costs that are generated. While the Three Strikes measure that is currently the law of the land in California may be defensible on some grounds not immediately apparent to this author, it is clearly not defensible on the basis that it makes the public safer by incapacitating dangerous offenders. The aging of the prison population may well be the most important challenge facing the California criminal justice system. This is a problem that predates the implementation of the state’s Three Strikes law. It is a natural consequence of mandatory minimum sentencing laws and other policies aiming to “get tough” by extending the length of incarceration. As observed in chapter 3, this “upping the ante” is one of the few ways left to appear “tough” to constituencies when incarceration comes to be such a widely used criminal sanction. It is up to the policy makers and citizens of California to decide whether the costs associated with supporting policies that have the consequence of aging the prison population are worth the benefits. It may be the case that Californians decide that these costs are worth bearing. If so, policy makers are obliged to examine other ways to maximize the level of dangerousness in the prison population, in order to better protect the citizenry. The three alternate policy scenarios presented in chapter 8 explore various ways to achieve greater average dangerousness in the incarcerated population. The first two of these scenarios accept the fact of an aging prison population and examines ways to focus the Three Strikes law more narrowly on particular types of offenders. Of the three variants of Three Strikes presented, it appears that the alternative that modifies the law such that eligibility for the harsher penalties provided for in the law is extended to only those offenders who have demonstrated a capacity for violence (either in the past or in the present offense) is the most effective from the standpoint of selective incapacitation. The last alternative explored in chapter 8 is that of implementing a program of geriatric release in California prisons. This option appears promising, both from the perspective of
Conclusion
155
maximizing dangerousness in the prison population, and from that of cost management in the criminal justice system, given the enormous costs engendered by incarcerating large numbers of elderly offenders. An important goal of this book is to promote a shift in thinking about policy making and framing policy choices. Aaron Wildavsky (1979) makes a simple claim in his classic work on policy analysis: “A promise underlies public policy: if the actions we recommend are undertaken, good (intended) consequences rather than bad (unintended) ones will actually come about” (Wildavsky 1979:35). It is hoped that if nothing else, the analyses presented here have demonstrated not only that policies need theories to guide them, but that the consequences of policies must be evaluated in terms of those theories. The haphazard, politically driven approach to criminal justice policy reform witnessed in California can be seen as largely responsible for the state of the state’s prisons, which are rapidly becoming what one observer has called crowded, “expensive old age homes for felons” (Zimbardo 1994:1). To paraphrase Wildavsky (1979:2), we need to learn what it is that the criminal justice system is better at doing, and what it is worse at doing. The experience of the past two decades indicates that the system is not very good at rehabilitation (or lacks the will to acquire this facility); the analyses reported in chapter 8 indicate that we could be better at the selective incapacitation of dangerous offenders than we currently are. The prospective analyses presented in chapter 8 are, in some ways, only a beginning. By demonstrating both the potential benefits that can result from fairly minor modifications to existing sentencing structures, as well as the potential for simulation analysis to allow for the prospective evaluation of proposed reforms, I hope to encourage further application of this methodology to explore the consequences of more drastic modifications to sentencing policy in California. These might include exploring some of the options that were just presented in combination (e.g., combining a program of geriatric release with that of restricted Three Strikes eligibility), or more fundamental reform, particularly in the area of penalty structures for drug offenses. As the retrospective analyses in chapter 7 demonstrate, the disturbing pattern of drug offenders increasingly occupying space in California prisons that might be used to house more dangerous offenders predates the introduction of Three Strikes by at least a decade. Simulation analysis opens the door for consideration of more “radical” policy ideas, at very little political or operational cost.
A WORD ON THE CRIMINOLOGIST’S PREDICAMENT Before the passage of California’s Three Strikes law, you would have been hardpressed to find a criminologist who believed that this law was a good idea. Yet,
156
Evaluating the Past, Choosing the Future
nearly three-quarters of California voters enthusiastically affirmed the measure that cemented the state legislature’s earlier action. Why? How could this happen? How is one to understand the disconnect between criminologists and the public they ostensibly serve? Criminologists often blame bad policy decisions on an irrational and vengeful public. Public opinion studies that examine the veracity of this assumption are at best, murky on this point (e.g., McCorkle 1993; Flanagan 1996; Applegate, Cullen, and Fisher 1997). Regardless of what the opinion research shows, we cannot place all the blame for the failure of the public to heed our advice on the public alone. We need to take seriously the idea that perhaps they do not listen to us because we do not speak to them. If we, as “experts” want to be taken seriously, we need to frame our pronouncements in ways that make sense to the public, and in ways that they can see as relevant to their own lives. As law professor Samuel H. Pillsbury (1995) exhorts: “we must resolve to engage the public, and not just ourselves” (Pillsbury 1995:317). One problem that Pillsbury identifies is the divergent time horizon of criminologists and the lay public. Citizens’ concerns are frequently short-term, and often focused on recent offenses that we, as experts on criminal behavior and criminal justice, are quick to point out don’t really happen all that frequently—so really, it is silly for them to worry about that. Criminologists, on the other hand, are more likely to take the long view, focusing on the outcomes of our interventions, or impacts on the size of prison populations, or on criminal justice expenditures. This split brings to mind the response of Harry L. Hopkins to fiscally conservative critics of Franklin Delano Roosevelt’s New Deal: “People don’t eat in the long run. They eat every day.” As Samuel Pillsbury (1995) persuasively argues, The criminal justice expert must recognize the inherently and specially emotional nature of criminal justice issues. . . . [E]xpert reliance on statistics to address the crime problem may signal a misunderstanding about human nature. For better or worse, most people connect with persons, not statistics. The Polly Klaas case touched a political nerve because its narrative was so compelling. . . . [I]n this sense, its statistical rarity is irrelevant. (Pillsbury 1995:320–321) If criminologists want their views on policy taken seriously, they need to start taking seriously the beliefs, opinions, and feelings of the people that these policies ostensibly benefit.
TOWARD A JURISPRUDENCE OF ANALYTIC REALISM At the beginning of this work, I stated that I wanted to frame this undertaking as a part of the analytic and realist schools of thought in criminology. A central
Conclusion
157
component of this endeavor has involved the re-orienting of strategies to evaluate the efficacy of policies purporting to further social defense goals via selective incapacitation away from traditional crime-counting approaches and toward the idea of the incapacitation of dangerous offenders. As discussed in chapter 6, operationalizing the construct of a “dangerous offender” required a certain degree of prioritizing with respect to the relative dangerousness of offender attributes, such as age, conviction offense, and demonstrated violence history. I advocate adoption of a similar approach in sentencing policy reform. It is time for California (and other states with similarly overburdened criminal justice systems) to begin to take steps to rationalize their criminal justice systems in order that these systems might function in the way that they are intended to, and the way in which the public expects them to. The analytic realist approach offers a framework that balances the realities of political necessity and the very real concerns about crime and the public safety. Policies that claim to enhance public safety by selectively incapacitating dangerous offenders need to accomplish two things. First, they need to make clear what a dangerous offender is, taking seriously the input of the public in formulating this definition. Second, they need to ensure that these dangerous offenders are actually the ones targeted for selective incapacitation. The formulation of a jurisprudence of analytic realism is not as immodest a goal as it seems. The roots for this undertaking have already been laid in such works as Zimring and Hawkins (1997) Crime Is Not the Problem and John Irwin and James Austin’s (1994) examination of the growth in prison populations in It’s About Time: America’s Imprisonment Binge (1994). These works highlight the absurdity of criminal justice policies that work to incarcerate large numbers of nonviolent drug offenders, particularly in an era when we enjoy, historically speaking, relatively low rates of crime. Innovation is needed in research, policy, and in the way in which criminologists understand their relation to both. While John Braithwaite’s (1989) dire pronouncements about the state of the field may overstate the case somewhat, his points are well-taken: The present state of criminology is one of abject failure in its own terms. We cannot say anything convincing to the community about the causes of crime; we cannot prescribe policies that will work to reduce crime; we cannot in all honesty say that societies spending more on criminological research get better criminal justice policies than those that spend little or nothing on criminology. Certainly we can say some important things about justice, but philosophers and jurists were making a good fist of those points before ever a criminological research establishment was created. (Braithwaite 1989:133) What criminologists can do is take up the realist project in order to use what we do know about crime and patterns of crime to aid in setting policies
158
Evaluating the Past, Choosing the Future
that have the potential to succeed, and to succeed in ways that will satisfy constituencies who want and deserve to be protected from dangerous offenders. It is important that criminologists learn to operate within the existing public policy discourse if our empirical work is to have any real impact on policy. Stanley Cohen (1988) admonishes that “radical [realist] criminology must make itself politically relevant by operating on the very same terrain that conservatives and technocrats have appropriated as their own. It cannot afford to risk the errors of the sixties by allowing itself to be marginalized” (Cohen 1988:19). I sincerely hope that this work can help to bridge the gaps between research, policy, and politics so that both can better achieve a common goal—the creation of a more just, livable, and safe society for all.
Technical Appendix A Data Sources and Estimation Procedures
The data used for validation of the simulation model presented in chapter 7 were obtained from various sources. Arrest data were provided by the California Department of Justice, Criminal Justice Statistics Center. Data on prison and parole entrances, exits, and daily population were provided by the Research Bureau of the California Department of Corrections. Detailed data on jail and probation populations were somewhat more difficult to obtain. Limited data on these populations are available from the California Department of Justice (1980–1997), and more specific values for each of the 450 relevant subgroups were estimated. Additionally, while the prison and parole data obtained were the most complete with respect to the six relevant offender characteristics (i.e., the five indicators of the dangerousness construct and race), other data sources contain less complete information. What follows is an account of the procedures used to develop the data in order to provide validation for the model, and details of the programs that were written to produce the four projected scenarios.
GENERAL FEATURES OF THE MODEL Berkeley Madonna1 (version 8.0) was used to create and estimate the simulation models. This software allows the user to represent a system as a series of differential and difference equations. Berkeley Madonna was developed by researchers in molecular and cellular biology at the University of California, Berkeley, for testing mathematical specifications of theoretical models of molecular, cellular,
159
160
Technical Appendix A
and developmental processes. I am unaware of other uses of this software in the social sciences. As discussed in chapters 6 and 7, the model consists of six population “states” and the pathways connecting them (e.g., the pathway by which offenders exit the state of being in jail at one time-step and enter prison at the next time-step). The simulation modeling software does not itself specify the metric of these time-steps. For this model, the incremental unit of time is one year, and validation data are annual data on the composition of populations and on transition probabilities. The transition probabilities are generally fixed, and represent the proportion of the “sending” population state that exits that state at each iteration of the model. Each transition probability is regulated over time by an informational quantity or multiplier. If a particular transition probability remains constant over time, the value of this multiplier is set to equal one. In the case of rising probabilities (e.g., the case of increasing drug arrests in the 1980s), the multipliers are set to values higher than one (and may be subject to “shocks,” delayed functions, or other stimuli to mimic changes over time). In the case of declining rates over time, multipliers have values lower than one. Generally, the transition probabilities for each population subgroup are not known. In most cases, a baseline transition probability (average, considering the entire population of both states) can be obtained, simply by dividing the size of the sending population by the size of the receiving population. In the absence of any theoretical justification to do otherwise, start values for transition probabilities for each of the four hundred fifty groups were assigned as this baseline, and altered in the course of model specification to reproduce system dynamics as represented by the validation data. For certain transition pathways, these base rates require modification, as in the case where the receiving state has more than one contributing state. For example, in the case of the transition rates that give rise to jail populations, this is done taking into account the relative proportions of sentenced and pretrial offenders in the population, and “prorating” the transition pathways to reflect these proportions. Different transition probabilities for each of the different subgroups are estimated in certain situations. If differing probabilities can be reasonably estimated from actual data using the division method, unique transition probabilities are applied for each of the population subgroups. The data on prison and parole populations are the most complete with respect to model attributes, and so many of the transition probabilities relating to these population states are estimated for each population subgroup for the transition rates between these two states, and for prison release rates. The second instance in which unique group transition probabilities are specified is when theoretical justification exists for doing so. This is most applicable in setting the transition
Technical Appendix A
161
probabilities for pretrial release (arrest to street) and pretrial detention. In general, nonwhite, male, younger, and violent offenders are more likely to be detained prior to adjudication than white, female, older, and nonviolent offenders (Irwin 1985; Steury and Frank 1990; Petee 1994).2
ARRESTED POPULATION As stated above, I received yearly data from 1979 to 1998 on adult felony arrests from the Special Requests Unit of the California Criminal Justice Statistics Center. These data were broken down by age, sex, race, and offense, but information about the criminal histories of arrested individuals was not available. In order to construct values for each of the population groups, I had to estimate the relative proportions of individuals with the various combinations of prior convictions and violence history. I estimated the proportions of arrested offenders in each category of criminal history (e.g., no prior convictions, 1 prior conviction, 1 violent prior, 1 prior conviction, 0 priors, etc.). These estimates were derived from the 1991 Survey of State Prison Inmates and were based on the proportions of convicted offenders in the various race/sex/age groupings having each of the five criminal history configurations. The same proportional estimates were applied within age/sex/race groupings, regardless of arresting offense. This decision was based on prior research that fails to show any meaningful evidence of specialization on the part of offenders, or any evidence of a patterned trajectory of offense transitions (e.g., Wolfgang, Figlio, and Sellin 1972; Wright and Rossi 1986). The model is structured such that the “arrested” population state clears on each iteration of the model. This means that 100% of the arrested population at t1 occupies a different state or leaves the model at t2. The general (noncriminal justice system involved) population is set up as a source/sink, as indicated here by the ellipse (see figure A1). In dynamic systems modeling, sources and sinks represent the boundaries of the model, in that they represent unlimited supplies of a particular input or resource (in this case, criminal justice system-involved individuals), and also serve as “absorbers” of resources as they exit the model (in this case, individuals leaving the criminal justice system by varied mechanisms such as acquittal, or completion of probation or other sentences). Because of this, the equation generating the arrested system state takes a slightly different form than that of the other population states. A new arrested population is generated at each time step by taking the values (number of persons in each population subgroup) of the arrested population at the first time-step and applying a multiplier to simulate growth or decline over time. The values of these multipliers for each population subgroup were
162
Technical Appendix A FIGURE A.1 Structural Model of the California Criminal Justice System
determined by fitting the output generated by the simulation model to the time-shapes of the actual data for each group over the twenty-year period that is the subject of the baseline model.
JAIL AND PROBATION POPULATIONS Since jail and probation are local rather than state functions, detailed statewide data on the composition of these populations are not available. This is true in part because of poor (or nonexistent) coordination among local agencies, but also in part due to the fact that simple demographic data on locally supervised populations are not collected by the agencies administering these functions.3 A similar approach was used to generate subgroup start values for both the jail and probation populations. Population totals for jail and probation were available broken down by sex only. These figures were further disaggregated into the 450 population subgroups based on the proportional representation of each age-sex-race combination in the total volume of arrests. From these groups, standard formulas were applied to parse out the subgroups on criminal history and current offense information. The proportions applied to the total jail and probation populations in order to generate values for each of the 450
Technical Appendix A
163
population subgroups were derived from the proportionate representation of each group in arrest volume. All offenders with two or more offenses were excluded from the probation population at the beginning of the simulation, on the assumption that a convicted offender with two prior felony convictions is extremely unlikely to receive a probation sentence (and after the advent of the 1994 Three Strikes law, for many offenders, a judge would be prohibited from handing down such a sentence even if he were so inclined).
“STREET” POPULATION The population state designated in the model as “street” contains those individuals who have entered the criminal justice system, and are awaiting adjudication but who are not detained in jail. Input values for this population were obtained by taking the proportion of the arrested population that is held over for adjudication (approximately 50 percent, that is, the cases in which official charges are filed) minus the proportion of the arrested population that is detained in jail prior to adjudication.
PRISON AND PAROLE POPULATIONS As stated earlier, relatively complete data on prison and parole populations (average daily population), entrances, and exits were obtained from the California Department of Corrections (CDC, or Department) . The data were complete with respect to age, sex, race, prior violence history, and conviction offense of offenders under the jurisdiction of the CDC. Some estimation was required, however, due to differences in the way offender criminal history information are tracked in the Department and the requirements of the model. The Department classifies offenders into one of four status categories: offenders entering prison or parole on a new offense who were not under the jurisdiction of the CDC at the time of offense commission are classified as A status. Offenders who are returned to custody for a new offense while still under Department supervision (e.g., on parole) are B status. C status individuals are those who are returned to custody on technical violations (noncriminal acts that nevertheless violate the conditions of parole, such as leaving the jurisdiction without permission), and D status individuals are those individuals returned to custody by who are in the process of mounting legal challenges to this return. For the purposes of this model, all of these offenders are equivalent insofar as they occupy a model population state (e.g., prison or probation). The difficulty arises in that offenders designated as A status have no recorded criminal history information in the CDC information system. This is misleading, as while they are treated as new
164
Technical Appendix A
entrants to the system, this is not necessarily the case. Offenders who have completed a term of parole or a prison sentence and are thus no longer under the supervision of the CDC would be classified as A status if arrested for a new crime. For this reason, the proportions of offenders with 0, 1, or 2 or more prior convictions and the proportions with or without a history of violence had to be estimated for the A status offenders. These estimates were based on the proportions of equivalent offenders (with respect to age, sex, race, and conviction offense) for which this information was provided (i.e., those offenders classified by the CDC as B, C, and D status). The equations used to generate the historical time series for the 450 prison and parole population subgroups were estimated in the same way as those for the arrested population, by using the software’s ocular least squares interface to fit the data series to the simulated series, by adjusting the rate modifiers appropriately.
THE PROJECTION ANALYSES Four different possible future scenarios were simulated to 2015 and reported in chapter 8. These include the likely consequences of the full implementation of Three Strikes, and two modifications of the Three Strikes future. The first of these modifications involves a narrowing of the “strike zone” to include as eligible only those offenders with a violent conviction offense and the requisite criminal history. The second Three Strikes scenario lies between these two extremes, and includes all offenders with a violent conviction offense and all offenders with a prior violent conviction. The Three Strikes scenarios were all estimated in a similar fashion. It will be recalled that California’s Three Strikes law includes provisions for both second- and third strike offenders. Second strike offenders are subject to a mandatory doubling of the presumptive sentence for the conviction strike, while third strike offenders are subject to twenty-five years to life, or three times the presumptive sentence, whichever is longer (with no parole eligibility for twenty-five years). The effects of these sentence length increases on the prison population were simulated by reducing the release rates from prison for eligible populations (offenders with one or two or more priors, with conviction offense and violence history varying according to the scenario). For second strike offenders, parole release rates were halved; for third strike offenders, both parole and direct-from-prison release rates were halved. This method preserves the scaling of parameters for different population subgroups. These changes were programmed using a delay function to take effect in 1999. The rationale for this is that most offenders subject to the provisions of the Three Strikes law would have been subject to some term of incarceration under existing law; the average length of prison stay in California is approximately three years (California Department of Jus-
Technical Appendix A
165
tice 1999). Additionally, given normal delays in processing and court backlog—which has been exacerbated since the implementation of Three Strikes—it is likely that there would be some delay in the manifestation of the effects on the criminal justice system. The fourth scenario estimated was the implementation of a geriatric release program in 2001. This was done by increasing the parole release rates to 90% for all offenders aged 55 and over, until these offenders comprised less than 0.5% of the prison population.
This page intentionally left blank.
Technical Appendix B Example of Simulation Model Code
What follows is the code used to simulate the baseline structural model in Berkeley Madonna for one population subgroup. As explained in chapter 7, the code is largely repetitive, with variability in start values and transition parameter estimates for different subpopulations. The following is provided to give the interested reader the opportunity to replicate the functional form of the simulation models presented in this book. METHOD RK4 STARTTIME 0 STOPTIME 31 DT 1.0 ARRESTEDwm1(t) ARRESTEDwm1(t dt) (pop2ar_wm1 ar2jail_wm1 ar2pop_wm1 ar2strt_wm1) * dt pop2ar_wm1 (NEWARRwm1* incrpop2ar_wm1) INIT ARRESTEDwm1 6452 NEWARRwm1 6452 bail_wm1 .599 INIT STREETwm1 3865 drop_rate_wm1 .373 detain_rate_wm1 .028 acquit_wm1 .10 convJL_wm1 .05 167
168
Technical Appendix B
convSTR_wm1 .10 INIT JAILwm1 665 INIT PROBATIONwm1 4,170 INIT PRISONwm1 70 INIT PAROLEwm1 24 endparole_wm1 .85 endpris_wm1 .15 endprob_wm1 .3470 incrpop2ar_wm1 DELAY (.6, 15, .85) incparo2pris_wm1 .15 incpris2paro_wm1 .3 incrar2jail_wm1 DELAY (1.7, 16, 1.03) incrar2pop_wm1 .9 incrar2strt_wm1 1.05 incrjail2pop_wm1 1.05 incrjail2pris_wm1 1 incrjl2prob_wm1 1 incrparo2pop_wm1 1 incrpris2pop_wm1 1 incrprob2jl_wm1 .9 incrprob2pop_wm1 .9 incrstrt2pris_wm1 .5 incrstrt2prob_wm1 DELAY (1.3, 14, .89) incstrt2jail_wm1 1 incstrt2pop_wm1 1 jailsentSTR_wm1 .07 jrelrate_wm1 .40; .05 {end sentence} parolrel_wm1 .7 probsentS_wm1 .19 probviolJL_wm1 .30 violparo_wm1 .6 violprJL_wm1 .3 ar2jail_wm1 (ARRESTEDwm1*detain_rate_wm1)*incrar2jail_wm1 ar2pop_wm1 (ARRESTEDwm1*drop_rate_wm1)*incrar2pop_wm1 ar2strt_wm1 (ARRESTEDwm1*bail_wm1)*incrar2strt_wm1 JAILwm1(t) JAILwm1(t dt) (ar2jail_wm1 strt2jail_wm1 prob2jl_wm1 jail2pop_wm1 jail2pris_wm1 jl2prob_wm1 ) * dt ar2jail_wm1 (ARRESTEDwm1*detain_rate_wm1)*incrar2jail_wm1
Technical Appendix B
169
strt2jail_wm1 (STREETwm1*jailsentSTR_wm1)*incstrt2jail_wm1 prob2jl_wm1 (PROBATIONwm1*probviolJL_wm1)*incrprob2jl_wm1 jail2pop_wm1 ( JAILwm1*jrelrate_wm1)*incrjail2pop_wm1 jail2pris_wm1 ( JAILwm1*convJL_wm1)*incrjal2pris_wm1 jl2prob_wm1 (PROBATIONwm1*violprJL_wm1)*incrjl2prob_wm1 PAROLEwm1(t) PAROLEwm1(t dt) (pris2paro_wm1 paro2pris_wm1 paro2pop_wm1) * dt pris2paro_wm1 (PRISONwm1*parolrel_wm1)*incpris2paro_wm1 paro2pris_wm1 (PAROLEwm1*violparo_wm1)*incparo2pris_wm1 paro2pop_wm1 (PAROLEwm1*endparole_wm1)*incrparo2pop_wm1 jail2pop_wm1 ( JAILwm1*jrelrate_wm1)*incrjail2pop_wm1 ar2pop_wm1 (ARRESTEDwm1*drop_rate_wm1)*incrar2pop_wm1 strt2pop_wm1 (STREETwm1*acquit_wm1)*incstrt2pop_wm1 pris2pop_wm1 (PRISONwm1*endpris_wm1)*incrpris2pop_wm1 prob2pop_wm1 (PROBATIONwm1*endprob_wm1)*incrprob2pop_wm1 paro2pop_wm1 (PAROLEwm1*endparole_wm1)*incrparo2pop_wm1 pop2ar_wm1 (NEWARRwm1)*incrpop2ar_wm1 PRISONwm1(t) PRISONwm1(t dt) (jail2pris_wm1 paro2pris_wm1 strt2pris_wm1 pris2paro_wm1 pris2pop_wm1) * dt jail2pris_wm1 ( JAILwm1*convJL_wm1)*incrjail2pris_wm1 paro2pris_wm1 (PAROLEwm1*violparo_wm1)*incparo2pris_wm1 strt2pris_wm1 (STREETwm1*convSTR_wm1)*incrstrt2pris_wm1 pris2paro_wm1 (PRISONwm1*parolrel_wm1)*incpris2paro_wm1 pris2pop_wm1 (PRISONwm1*endpris_wm1)*incrpris2pop_wm1 PROBATIONwm1(t) PROBATIONwm1(t dt) (jl2prob_wm1 strt2prob_wm1 prob2pop_wm1 prob2jl_wm1) * dt jl2prob_wm1 (PROBATIONwm1*violprJL_wm1)*incrjl2prob_wm1 strt2prob_wm1 (STREETwm1*probsentS_wm1)*incrstrt2prob_wm1 prob2pop_wm1 (PROBATIONwm1*endprob_wm1)*incrprob2pop_wm1 prob2jl_wm1 (PROBATIONwm1*probviolJL_wm1)*incrprob2jl_wm1 STREETwm1(t) STREETwm1(t dt) (ar2strt_wm1 strt2pop_wm1 strt2jail_wm1 strt2prob_wm1 strt2pris_wm1) * dt ar2strt_wm1 (ARRESTEDwm1*bail_wm1)*incrar2strt_wm1 strt2pop_wm1 (STREETwm1*acquit_wm1)*incstrt2pop_wm1 strt2jail_wm1 (STREETwm1*jailsentSTR_wm1)*incstrt2jail_wm1 strt2prob_wm1 (STREETwm1*probsentS_wm1)*incrstrt2prob_wm1 strt2pris_wm1 (STREETwm1*convSTR_wm1)*incrstrt2pris_wm1
This page intentionally left blank.
NOTES
NOTES TO CHAPTER 1 1. Analysis of the complex political and psychological processes that govern the acceptance or rejection of a particular paradigm is beyond my scope of inquiry. My task is to identify criminal justice paradigms as they have manifested themselves in criminal justice sentencing policy as a precursor to the empirical objective of the study, the evaluation of the impacts of policy choices on prison populations. 2. This task has been quite admirably undertaken by a number of contemporary scholars, including Nigel Walker (1991); Andrew von Hirsch (1976, 1985); Herbert L. Packer (1968); Norval Morris (1974); Franklin E. Zimring and Gordon Hawkins (1995); Thomas Mathiesen (1990); and H.L.A. Hart (1968), to name just a few. 3. Another important difference is the period of study—many of the large-scale research projects on sentencing policy reform (e.g., Clear 1994; Wichayara 1995; Zimring and Hawkins 1995) only analyze data up through the early 1990s, well before the implementation of its most dramatic sentencing reforms, such as Three Strikes. 4. It is true that the California penal system underwent a great deal of programmatic and structural reform under the direction of Richard A. McGee, the first director of the state’s Department of Corrections (Glaser 1995). However, these reforms did not have any real paradigmatic significance, in that all of McGee’s reforms were directed toward the goal of rehabilitating and reintegrating offenders into society. 5. A more subtle point here is that dangerousness is ultimately a probability—to say that an individual is more dangerous than another is to say that one individual has a higher probability of being dangerous than another. The same point (although expressed somewhat more laboriously) holds for populations: what we are really claiming is that one population is more likely than another to have a greater or lesser number of individuals with a greater or lesser probability of behaving dangerously in the future. 6. Dynamic systems models have also been used to simulate various aspects of illicit drug use. Some of these models focus on large-scale drug-distribution networks (Childress 1994a, 1994b; Dombey-Moore, Resetar, and Childress 1994), while others model populations of drug users (Hanneman and Jacobsen 1992).
171
172
Notes to Chapter 2
7. This perspective derives, in large part, from Thomas C. Schelling’s (1978) observation that any number of different processes may, in fact, lead to similar outcomes. 8. Explaining these differences, while a worthwhile endeavor, is beyond the scope of this analysis. Many researchers do attempt to explain them (e.g., Irwin 1985; Meyers 1987; Bridges and Steen 1998), and the reader interested in the reasons such differences exist is referred to these authors. For my purposes, these differences are merely noted as “social facts” and incorporated into the modeling strategy in an effort to reproduce system dynamics as accurately as possible. 9. I make this explicit disclaimer in a perhaps futile attempt to safeguard against incorrect interpretations of the modeling strategy I employ. The most sensitive attribute essential to this modeling process is that of race: while race is related (with varying degrees of strength) to transition probabilities in the criminal justice system, race is not incorporated into the measure of dangerousness. This issue is discussed further in chapter 5. 10. This period spans the earliest and latest dates for which detailed electronic validation data are available from California state criminal justice agencies. 11. Evaluating changes in the average level of dangerousness in correctional populations that result from particular changes in sentencing practices is an informative exercise even if such reforms were not explicitly concerned with incapacitating dangerous offenders. Understanding the impact of particular policy changes on the distribution of dangerous offenders in the population is useful, insofar as crime reduction (by whatever specific mechanism, e.g., deterrence, incapacitation) is considered as an objective of criminal sentencing at all. 12. I would also term this tradition the “social constructionist” school.
NOTES TO CHAPTER 2 1. Spierenburg (1995) points out that although prisons are at the forefront of most people’s perceptions of the criminal justice system, “the most common judicial sanction is a fine (usually for violating traffic rules)” (Spierenburg 1995:61). 2. There is precedent for this usage of the paradigm framework; Bertram et al.’s (1996) work on drug policy contains an analysis of paradigms in American criminal justice. 3. Interestingly, although the modern prison was essentially an innovation of the rehabilitative paradigm, successive penal paradigms have done little to challenge its dominance. Explaining the persistence of the prison is beyond the scope of this analysis, but see Garland (1990) for a discussion of this issue. 4. In practical terms, the gulf between the “social defense” and “desert” paradigms is not so large as some zealous proponents of each of these positions would have it. Undoubtedly, supporters of both of these paradigms would generally agree as to who ought to be incarcerated for long periods of time (e.g., an offender convicted of rape vs. an offender convicted of loitering).
Notes to Chapter 2
173
5. Although retribution fits most squarely into the designation of “punishment for its own sake,” the potential social benefit of retribution cannot be denied (e.g., Durkheim 1984). 6. The death penalty is a special case, in that execution can be seen both the ultimate deterrent and the ultimate form of incapacitation. However, incarceration has figured more centrally in the incapacitative enterprise in the late twentieth century. The most convincing justifications for the use of capital punishment tend to focus on the retributive aspects of the death penalty (Packer 1968; Walker 1991). 7. These two examples of Roman law illustrate Thorsten Sellin’s (1976) thesis involving class differences in punishments as delineated in ancient law. The Twelve Tables show a marked bias in this regard. Offenses that could only be committed by the relatively well-off (e.g., setting one’s herd of cattle to graze another’s pasture, which presupposes ownership of a herd of cattle) tend to be addressed in restitutive terms, while offenses to which the less-privileged have equal access tend to be punished much more severely. Law XII is another prime example: “Anyone who gives false testimony shall be hurled from the Tarpeian Rock” (Scott 1932:71). 8. Morris and Rothman (1995) contend that the trial process itself presupposes the existence of a house of detention. 9. Friedman (1993) notes that this period also marks the emergence of distinct criminal and civil systems of justice; the feudal church was seeking to reduce the incidence of “blood feuds” between perpetrators and victims of wrongdoing (see also Hibbert 1963). 10. The Norman conquest also marks the beginning of the process of separating church law and jurisdiction from that of the Crown. While the church still had a considerable scope of authority, its authority becomes differentiated from that of the civil justice system (Kuntz 1988; Briggs et al. 1996). 11. Langbein (1998) notes that the linguistic transformation takes place in Germany, where “the Dutch tuchthuis became in German the zuchthaus, a word which lost the meaning of “workhouse” for vagabonds and petty offenders and acquired the modern sense of ‘prison’ or ‘penitentiary’ for serious offenders” (Langbein 1998:13). 12. Indeed, the linkages between poverty, vagrancy, and punishment are numerous (see Garland 1985 for a more thorough treatment of these ideas). Both transportation and galley slavery also had elements of “discipline” as a way of dealing with poverty and its consequences—apart from their function as criminal punishments (Weidenhofer 1973; McGowen 1995; Spierenburg 1995). 13. Friedman (1993) contends that criminal justice in colonial American did not quite approach the level of brutality extant in medieval Europe, relying less on execution and more heavily on lesser corporal punishments such as public whipping and mutilation. However, if comparisons are limited only to the relevant period (i.e., the seventeenth century and later), the argument becomes less convincing, in that the origins of the ideological shift that produced the rehabilitative paradigm were already beginning to be seen in the gradual substitution of more private forms for punishment for public ones that begins in eighteenth-century Europe (Foucault 1977).
174
Notes to Chapter 2
14. Although Michel Foucault’s analysis focuses on the technologies of punishment in Europe, developments in the United States follow a similar pattern. 15. Indeed, this is an extreme simplification of Elias’s (1982 [1939]) thesis. Elias proposed the civilizing process as a multilayered and self-reinforcing phenomenon, which began in the development of manners in the court societies of Europe and that he argues effected changes in the consciousness of individuals. This led to the development of interactional norms prescribing ever-greater levels of self-restraint, which in turn leads to the elaboration of institutions (e.g., the state) and institutional norms regulating expression and interaction among individuals. This process as explicated is selfreinforcing and self-perpetuating, in that social institutions are presumed to have a significant impact on the individual’s consciousness: A continuous, uniform pressure is exerted on individual life by the physical violence stored behind the scenes of everyday life [a consequence of the monopolization of violence by the modern state], a pressure totally familiar and hardly perceived, conduct and drive economy having been adjusted from earliest youth to this social structure. It is, in fact, the whole social mould, the code of conduct that changes, and accordingly with it changes, as has been said before, not only this or that specific form of conduct, but its whole pattern, the whole structure of the way individuals steer themselves. (Elias 1982 [1939]:239). 16. Helmut Kuzmics (1988) argues this point quite adroitly in addressing critics of Elias who argue that the Nazi death camps cannot be explained by the civilizing process thesis; Kuzmics counters that “the ritualized and bureaucratic character of this monstrous administration of murder makes clear that the motivation of those who performed the murders was comparatively irrelevant” (Kuzmics 1988:157). Although barbarous and brutal, the Nazi camps were indeed private places where the torture and killing of “inmates” took place outside of public view. 17. Philips (1983) locates this shift within an increasing focus on individualism in Western thought; this focus on the uniqueness of individuals also contributes to the ascendance of the treatment ideal embodied in the rehabilitative paradigm. 18. Rothman’s argument is consistent with the notion of the civilizing process, that was just discussed. 19. Several scholars have noted with interest the parallel development of the modern penitentiary in America and England; these scholars are generally in agreement that England’s intellectual contribution to penal developments across the Atlantic was minimal (e.g., Hibbert 1963; DeWolf 1975; Rothman 1995). 20. Positivist criminology logically led its proponents down two seemingly different but conceptually interrelated paths: the first of these, largely embraced by American “New Penologists,” was that of Quetelet’s (1842) original position—that the causes of crime were to be found within society. The biological determinism usually associated with Cesare Lombroso was actually first articulated by Quetelet in the 1840s (Beirne 1993:90). While the two paths may lead to different response modalities (e.g., treatment
Notes to Chapter 2
175
vs. eugenic population control strategies), they are profoundly compatible with the essence of Quetelet’s positivism, in that the cause of the behavior is, in both instances, rooted in something outside of the will or choice of the offender. 21. Although the differences were relatively minor, the superiority of one system over another was hotly debated throughout the nineteenth century. The chief advantage of the “silent” system was the lower cost it entailed to implement; detractors argued that although less expensive, the silent system produced only “temporary obedience,” while the separate system was capable of producing true change in the offender through penitent reflection (McKelvey 1936; Rothman 1971; Kuntz 1988; Friedman 1993; McGowen 1995). In the end, New York’s “silent system” emerged as the clear winner, doubtless as a consequence of the cost advantages (McKelvey 1936; Rotman 1990). 22. In fact, Alexander Pisciotta’s (1983) analysis of records of the Elmira reformatory during the tenure (1876–1900) of the venerated “penal scientist” Zebulon R. Brockway reveals that at least this particular institution was not a progressive and benevolent institution where inmates were cared for in a constructive and humane manner, but rather a place where beatings, threats, hard labor, and excessive discipline were the norm. 23. McKelvey (1936) points out that although the rehabilitative paradigm was fairly well entrenched in the United States by the 1860s, the ideas spread unevenly across the nation; wretched prison conditions persisted in many parts of the country, particularly in parts of New England and in some southern states (McKelvey 1936:19–20). 24. The position of Martinson and his colleagues was not quite this simplistic. They found no significant differences in recidivism rates between groups of offenders exposed to different programs—which would, strictly speaking, be more appropriately expressed as “Nothing works much better than anything else.” But “nothing works” was the slogan that caught on—and is more or less true to the essence of the findings. 25. This is an essential feature of paradigm shifts. Consider the following metaphor: if one declares that a machine isn’t working, there is a clear implication that it can be fixed, and is perhaps capable of working at a later time. It is quite a different scenario to claim that this particular machine is entirely the wrong instrument for the job at hand—which would require a new conceptualization of the job as well as of the tools and techniques needed to accomplish it. 26. While this period was characterized by a great deal of consensus, there was a fairly vocal minority (drawn primarily from the liberal side of the political spectrum) that questioned the wisdom of abandoning discretion entirely. The views of some of these dissenters are presented in a monograph, Determinate Sentencing, the proceedings of conference held at Boalt Hall Law School in Berkeley in 1977. 27. Rothman (1983) also asserts that the shift toward determinate sentencing is expressive of a larger social trend—the rejection of “experts” of all kinds, the mistrust of discretion in general, and the changing relationship Americans had with authority during the period of the 1960s and 1970s, as a result of, among other things, the Watergate scandal and the Vietnam War. Francis A. Allen (1981) makes a similar argument.
176
Notes to Chapter 2
28. Interestingly, Langbein (1998) notes that determinate sentences in the early modern period were underlied by a rehabilitative justification, in that these sentences emerged in a context in which the only alternatives were death and life sentences (Langbein 1998:13–14). 29. Others include Zimring and Hawkins (1973); Gibbs (1975); and Blumstein et al. (1978). 30. Twenty years later, Nagin’s (1998) viewpoint is much the same: [T]he accumulated evidence on deterrence leads me to conclude that the criminal justice system exerts a substantial deterrent effect. That said, it is also my view that this conclusion is of limited value in formulating policy. Policy options to prevent crime generally involve targeted and incremental changes. Thus, for policy makers the issue is not whether the criminal justice system in its totality prevents crime but whether a specific policy, grafted onto the existing structure, materially adds to the preventive effect. . . . While it is my view that the evidence points to the entire enterprise as having a substantial impact, predicting the timing, duration, and magnitude of the impact of incremental adjustments in enforcement and penalties remains largely beyond our reach. (Nagin 1998:331/338) 31. Minnesota is a notable exception; that state’s sentencing commission explicitly required prison capacity limits be taken into account in making sentencing decisions. 32. Analyses of citations in the criminology literature conducted by Ellen G. Cohn and David P. Farrington (1994, 1998) provide evidence of the prominence of the authors discussed. 33. Gabriel Tarde’s (1903) influence on American criminology has been more widespread than is generally recognized. Tarde’s “imitation” thesis of the etiology of criminal behavior is the foundation for later social-psychological theories of crime causation such as Edwin H. Sutherland’s differential association (1947) and anomie theory (Merton 1938). 34. While both of these strategies invoke the social defense objective of incapacitation, collective incapacitation is much more consistent with the retributive paradigm than is selective incapacitation, in that the primary factor considered in sentencing is the past and not the future behavior of the offender. 35. Rehabilitation still has its stubborn advocates (e.g., Cullen and Gilbert 1982; Shichor 1987; Rotman 1990; Palmer 1992, 1994), but evidence of their influence in contemporary sentencing reform is essentially nonexistent. 36. Interestingly, the rise of actuarial criminology, like the reemergence of (selective) incapacitation, has also been identified as a result of the failure of prior paradigms in criminology (Young 1986; Braithwaite 1989). 37. Although they do not say so explicitly, this is obviously a nod to the “original” New Penology of the late eighteenth century; they (wisely, in my view) apparently
Notes to Chapter 3
177
decided to refrain from making this explicit by terming their formulation the “new new penology.” 38. I resist the urge to label the new paradigm neo-selective incapacitation. 39. Timothy J. Flanagan (1996) notes that “the apparent ascendance of punishment as American’s choice for the chief goal of prisons may be a function of the fact that pollsters have not provided an alternative to the punishment [and rehabilitation] response[s]” (Flanagan 1996:82). In other words, it is not clear from surveys that find that respondents choose “punishment” over “rehabilitation” whether respondents are registering support for punishment, or merely rejecting rehabilitation.
NOTES TO CHAPTER 3 1. This was, of course, a feature of every state’s indeterminate sentencing structure. 2. The fact that these views are often deliberately manipulated by political elites is acknowledged, but is not analyzed in this context. The derivation of popular opinions about crime and punishment is a complex subject and will not be tackled in depth here. For our purposes, popular views on crime and the ways in which they have changed over time in California are treated as social facts, and employed in the analysis of the relationship between criminal justice paradigms and sentencing policy. 3. Approximately half of the states have some form of the ballot initiative process. Only a handful of these states frequently use the initiative; these include Oregon, Washington, and California. Western states tend to be overrepresented both groups (i.e., those that have the initiative process and those that use it frequently). 4. The Three Strikes statute has a complicated history. It was initially proposed to the state assembly by Bill Jones and Jim Costa (both Democrats from Fresno, Mike Reynolds’s home district), and defeated in committee. The measure was later passed into law by the legislature in March 1994, and a duplicate measure was affirmed by the electorate in November 1994. This is discussed further below. 5. It should be noted that the majority of criminal justice ballot initiatives during the period of interest were initiated in the legislature, as opposed to grassroots civic groups. However, the support of the people is evident in the voting majorities approving the various measures. 6. The exact number of signatures required varies depending on the type of measure being considered. In order to qualify an initiative statute, a number of valid signatures not less than 5% of the number voting in the last gubernatorial election must be gathered; for a measure seeking to amend the state constitution, the number of valid signatures required is 8% of the number voting in the last gubernatorial election (Schrag 1998:192–193). Peter Shrag also notes that in most cases, a large margin of signatures above this minimum are actually collected, in order to assure qualification in the face of stringent verification requirements.
178
Notes to Chapter 3
7. The Attorney General’s office, the initiator of the measure, noted that the Legislative Analyst’s Office determined that implementing the facial implant system would result in unknown costs, likely several million dollars per year. However, given Californians’ willingness to approve other, more expensive measures in the same period (such as Three Strikes), it is unlikely that this was a determining factor in the measure’s failure. 8. Actually, California’s first “prison” was a hulk named Wabau, moored in San Francisco Bay; convicts confined on the ship worked in a quarry on Angel Island (Lamott 1961). 9. One consequence of this nativism is the almost complete invisibility of retributive sentiment in early policy statements concerning the objectives of the state prison; as Shelley Bookspan (1991) puts it, “the prison only incidentally served even retributive goals because . . . aliens were not seen as treasuring the liberty of which incarceration deprived them” (Bookspan 1991:xviiii). 10. The importance of discretionary parole as a regulating mechanism becomes increasingly evident upon examination of the consequences of its elimination under determinate and mandatory sentencing. 11. This was also true elsewhere in the West (see, for example, Woznicki 1968:37–41). 12. So strong was this commitment that he announced his retirement in 1967 from the state cabinet post to which he had been appointed in 1961, after the election of Governor Ronald Reagan, whose campaign platform relied heavily on promises to mete out harsher penalties to convicted criminals (Glaser 1995:25). 13. In this and in subsequent discussions of ballot propositions, the reference is the state ballot pamphlet for the election in question. 14. The provisions of the Boatwright amendment generally increased penalties, chiefly by removing limitations on sentence enhancements that could be added on to the base term. 15. The penalty enhancement structure provided for a three-year add-on for each violent prior conviction resulting in incarceration if the current conviction offense was also violent, or a one-year add-on for each prior incarceration if the conviction offense was not violent. The existence of an enhancement for prior convictions speaks to the persuasiveness of the selective incapacitation idea in California; the structure of the enhancement schedule articulates this interest even more fully, given the wide disparities in available enhancements depending on whether or not the offender has a history of violent conduct (violence will later be considered as a central component of dangerousness). 16. In addition, Proposition 7 revised the statute dealing with the consideration of mitigation and aggravation in such a way that required judges and juries to give greater weight to aggravating than to mitigating circumstances in determining sentence. 17. Marc Klaas ultimately became one of the most vocal opponents of Proposition 184. The rhetoric of the organized opposition did not argue with the paradigmatic
Notes to Chapter 4
179
objective—selective incapacitation—of the measure, but rather with the ability of the Three Strikes law to achieve that objective. Klaas’s was the first name listed below the statement against the proposition printed in the ballot pamphlet for the November 1994 election, which argued that “if Proposition 184 passes, our prison system will be bloated with non-violent offenders serving life terms” (G94). 18. This feature of the law drew some challenges. Many sentence enhancements result from the existence of a prior conviction; the challengers asserted that it constituted a violation of the double jeopardy principle to double the presumptive sentence based on the prior conviction “strike,” and then to also add the sentence enhancement due to the same prior conviction. The courts disagreed on two separate occasions (People v. Ramirez [1995]; People v. Jackson [1995]). 19. Indeed, the first “Three Strikes” style statute in the nation was enacted in Virginia in 1796 (Ziegler and del Carmen 1996). 20. These consist of the Part 1 felonies of murder or nonnegligent manslaughter, rape, robbery, and aggravated assault (Ditton and Wilson 1999). 21. This narrative continues with the statement “according to Cohen, 46 percent of the state’s probationers commit new crimes” (Schrag 1998:100).
NOTES TO CHAPTER 4 1. This does not, of course, prevent the offender from preying upon his incarcerated fellows; however, this is rarely considered in discussions of the incapacitative benefits of imprisonment. 2. The concordance of the proportion of high-rate offenders in incarcerated and cohort samples (the approximately 20% figure) raises the interesting possibility that inmate samples may actually be representative of the general population of criminal offenders in society. It is unlikely, however, that such a proposition will ever be definitively resolved. 3. Chaiken and Chaiken’s interpretation of their findings from the Rand data are sometimes contradictory and confusing. At times they indicate a belief that there is a lifecourse progression of seriousness on the part of offenders (Chaiken and Chaiken 205–206); however, since they note that the violent predators tend, on average, to be younger than other offenders (Chaiken and Chaiken 209), on a closer examination it seems that this impression may derive from the Chaikens’ presentation of the data in a hierarchical fashion. It seems that the researchers are engaging in an imputation of the relationship of groups of criminal offenders to one another to the relations that exist between individual offenders (a simple ecological fallacy). Similarly, Chaiken and Chaiken emphasize, at various points, that several of the identified types of criminals do appear to be highly specialized in their modes of criminal behavior; however, the “violent predators,” who commit the greatest proportion of crimes in the sample, do not show such evidence of specialization. I find James D. Wright and Peter H. Rossi’s (1986) interpretation of
180
Notes to Chapter 4
similar findings in their research (Wright and Rossi 1986:76; see below) a much more satisfying description of the pattern that presents in the Rand data. 4. What is actually being asserted in a statistical prediction is not that “individual X has a 60% probability of exhibiting behavior Y,” but rather that “in the past 60% of people who are like individual X have exhibited behavior Y.” 5. These tables underscore Shah’s (1981) assertion that exceedingly rare behaviors (e.g., violence or high-rate offending) do have fairly high base rates in certain populations, such as incarcerated felons. However, the dismal accuracy level in identifying high-rate offenders even in the presence of relatively high base rates should explain why no one has proposed that we try to prospectively identify violent individuals in the general population. While there has been renewed interest in the role of physiological and genetic factors in recent years (e.g., Kandel and Mednick 1991; Fishbein 1996), no one has yet articulated a specific plan for widespread bioscreening to root out violent individuals. Rather, prescriptive statements arising from this research are extremely vague, as in Diana Fishbein’s (1996) concluding statements that Instead of waiting until a vulnerable child becomes old enough to incarcerate, perhaps early assistance will enable us to avoid the personal and financial expense of criminal justice system involvement. There is little evidence that present tactics are effective; thus, we need to move forward into an era of early intervention and compassionate treatment that genetic research may advance. (Fishbein 1996:93) 6. As von Hirsch (1985) has also noted (von Hirsch 1985:110–114), Greenwood and Abrahamse (1982) report a false-positive rate of only 4% (Greenwood and Abrahamse 1982:59); they arrive at this figure by identifying only the most extreme errors in prediction, those predicted to be high-rate who are actually low-rate. This disingenuous reporting strategy conceals the fact that even the medium-rate offenders erroneously predicted to be high-rate (which brings the false-positive rate up to 48%) would be subject to the same consequences of incorrect predictions (namely, lengthy incarceration) as the false positives that are acknowledged in the limited definition used by the report’s authors. 7. My discussion focuses on the problem of false positives as I believe that these present a serious ethical challenge that is uniquely enhanced by selective incapacitation schemes. This is due in large part to the relative rarity of high-rate offenders, which increases the difficulty in making accurate predictions. False negatives, or offenders not identified as likely to be dangerous who turn out to be, are a threat to the public safety that exists in any sanctioning system that incorporates predictive judgment. However, an overreliance on a particular “scientific” scheme for assessing dangerousness might well exacerbate the problem of false negatives. 8. Running away and “incorrigibility” are examples of status offenses. They are so called because although these behaviors are punishable by the juvenile justice system, they are not technically criminal acts, and they are behaviors that are only sanctioned when committed by a minor child (i.e., they are defined as criminal due solely to the status of the acting individual).
Notes to Chapter 5
181
9. Additionally, in many states, prior convictions are already taken into account in determining the appropriate sentence for offenders. 10. I refer the reader to Avi-Itzhak and Shinnar (1973), Shinnar and Shinnar (1975), and Greenwood with Abrahamse (1982:74–77) for the computational details of estimating incapacitation effects. My purpose is merely to offer the historical background of the idea of selective incapacitation. Estimates of incapacitation effects are underlied by the assumption that high-rate offenders can be accurately and reliably identified—a premise that I believe to be untenable. 11. Greenwood and Abrahamse did not consider career length at all in the estimates of incapacitation effects provided in their 1982 report. 12. Zimring and Hawkins (1995:chapter 7) contains an excellent critical assessment of the validity of engaging in speculation about the “social costs” of crime versus those of punishment.
NOTES TO CHAPTER 5 1. One need only consider Durkheim’s thoughts on this matter to explain why this would be so; the intentional violation of norms is inherently more damaging to the collective morality (in that it entails an explicit rejection of that morality) than is a violation that comes about as a result of ignorance. 2. There is room for dispute on this point. One line of attack might cite the earlier discussion of the socially determined nature of dangerousness, and claim that once a particular risk is deemed intolerable, then all circumstances/persons/situations that fall above that threshold of risk are categorically dangerous. This seems to be the approach taken by Norval Morris and Marc Miller (1985). Later in this chapter I hope to demonstrate the flaws of this approach when considering the future behavior of individuals. 3. It is ironic that there have been so many attempts to predict something that is, by its very nature, unpredictable. Irony aside, it is not at all surprising that such attempts have enjoyed such little success. 4. This is the case regardless of the basis for our decision that the risk is unacceptable, that is, whether it is the magnitude of the probability of harm or nature if the harm expected. 5. Schrager and Short (1980) note that even when white-collar crimes are prosecuted and sanctioned, the primary focus is on the culpability of the offender, rather than on the outcome of the offense, something that is of primary importance in the processing of “ordinary” street crimes (e.g., murder vs. attempted murder). 6. A notable exception is the application of the dangerous label to the criminally insane, whose inability to form intent is their primary defining characteristic. 7. In a similar vein, they argue for the inclusion of juvenile criminal history measures in consideration of the sentencing dispositions of dangerous offenders: “The juvenile
182
Notes to Chapter 6
offenses of someone who continues committing crimes as an adult have a different status than if he had stopped. They look like early evidence of a blameworthy character rather than youthful indiscretions” (Moore et al. 1984:60). I find this line of argument exceedingly obtuse, in that it is grounded in the notion that actions taken at some later point in time change the character of prior behaviors. A logical extension of this would necessitate the excuse (or at least the redefinition) of the behavior of a serial killer if, at some point after killing a number of innocent young women, another young woman violently attacked him without provocation. Could his prior behavior then be construed as self-defense? 8. The determination of an individual or group as dangerous is a self-reinforcing process. The label of dangerousness serves a separating function: a person who is dangerous is somehow different from us. Therefore, labeling someone as dangerous increases the social distance between him and ourselves, thus intensifying this quality of unknowability. 9. One could also include the massive federal programs devoted to the study of criminal careers. I include in this discussion only projects with an explicit focus on dangerous offenders. 10. This kind of prediction can be considered “pseudoclinical.” Since the jurors do not draw on a specific body of training or knowledge, it cannot properly be called “clinical,” but it is analogous to clinical prediction in terms of the method. In addition, there is some evidence that there are no significant differences between judgments of dangerousness made by laypersons and professional clinicians ( Jackson 1988). 11. The commutations resulted from a variety of circumstances; see Marquart et al. (1989) for details.
NOTES TO CHAPTER 6 1. These differences are largely explained by the inclusiveness of the California law. California’s “strike zone” encompasses a wide range of felony offenses that are excluded from other states’ statutes, such as drug violations and burglary (Kempinen 1997; Austin et al. 1999). In other states, Three Strikes laws have largely replicated existing penalty structures, so their primary impact has been merely symbolic. 2. This strategy was first suggested by Gordon (1977). 3. It should be noted that although the proposed measure of dangerousness is used in a retrospective and evaluative sense, there is an implicit predictive component. This is because dangerousness is inherently a potentiality—the likelihood that an individual will be dangerous. 4. Blumstein (1982:1262) points out, in considering the race-dangerousness issue, the extreme sex differences in incarceration as well as in all measures of criminal offending. The inclusion of sex in the dangerousness construct underscores the rationale for excluding race; there are no similarly dramatic and overarching patterns in other factors, such as SES, that might plausibly influence dangerousness that vary solely by sex.
Notes to Chapter 6
183
5. Additionally, it is hoped that the use of conviction data will serve to make the model more “portable”—that is, more easily adapted for use in evaluating sentencing policy in other jurisdictions, where arrest reporting procedures may be different than in California. 6. The literature on vulnerability and fear is particularly instructive in this respect. It has consistently been shown that individuals who are more physically vulnerable (e.g., women and the elderly) fear crime more than do those who are less physically vulnerable (Warr 1984; Hale 1996). This can arguably be construed to correspond more to fear of violent crime than to property crime. 7. This relates directly to Ferraro’s (1995) contention that what really matters is the perception of risk, which may depend on a host of factors (e.g., media exposure and vicarious experiences via friends and acquaintances). 8. This analysis also reported that two-thirds of these low-level drug offenders received mandatory minimum sentences; the authors further reported that these offenders comprised over 20 percent of all federal prisoners. 9. The use of the term “states” follows the approach of Jay Forrester (1961, 1968, 1969a, 1969b, 1973) as explicated in Hanneman (1988). This usage literally corresponds to “states of being”, in that an individual can occupy only one state at any one point in time (e.g., an individual can be in prison or jail, but not both). 10. The criminal justice system as here conceived is characterized by an extremely high level of connectivity. For this reason, graphic representations are of extremely limited utility. I offer this figure merely as a schematic of the process that is being described. 11. Although race is not directly relevant to the central concern of the analysis, it will be instructive to estimate the impacts of sentencing reforms with respect to the racial composition of California’s prison, particularly in light of the fact that one goal of determinate sentencing reform was to reduce racial disparity in sentencing, as well as recent evidence that African-Americans are disproportionately affected by California’s Three Strikes law (Davis, Estes, and Schiraldi 1996). 12. These data are discussed in Technical Appendix A. 13. A common way of circumventing problems of identification in applications of structural equation modeling involves the imposition of a number of simplifying assumptions. I believe that the overburdened condition of California’s criminal justice system has come about in large part due to a failure of researchers, politicians, and practitioners to attempt to anticipate the consequences of policy actions in light of the complexity of the system. The skeptical reader may claim that I am defending a fictional methodology by highlighting the fictional qualities of another. However, although simulation modeling is indeed “simulated” and therefore does bear some resemblance to fiction, the first stage of the modeling process entails replicating the existing system. Although the model will be simulated (as necessitated by the high degree of complexity), it is constructed with explicit reference to the available data and is therefore grounded in empirical reality, rather than in assumptions.
184
Notes to Chapter 7
NOTES TO CHAPTER 7 1. In dynamic systems modeling, a single iteration of the model is conceived as a “time-step.” It falls to the individual analyst to specify (or choose not to) the material definition of those time-steps. In this analysis, annual data are employed to validate the model, so each time-step is defined as one year. 2. Berkeley Madonna also allows for the specification of more complex functional forms. The one most commonly used here is a step or delay function, which allows for the modeling of nonmonotonic functions. The most common of these forms is characterized by a rate of increase at time (t) through (t X), and then by a greater rate of increase for the time period after (t X). 3. The specification of the model in its entirety runs to over a thousand pages of code. For this reason, the model code will not be reproduced here. This example is provided to demonstrate the general form of the code. 4. No actual data are available for the “street” population—those awaiting adjudication but not detained in jail—so the simulation series is not validated and is thus omitted here. 5. Due to the way in which the equations that compile the subpopulations classified by dangerousness are programmed, levels of dangerousness for the first year of data (1979) cannot be calculated; therefore, the following analyses examine the period from 1980 to 1998. 6. This is arrived at by dividing the higher value (7.34) by the difference between the highest and lowest points in the series (.64). This is the logic that will be employed throughout this discussion, unless otherwise noted. The denominator of the fraction used to derive the percent changes will always be the point (highest or lowest) that comes first in time. For example, if the dangerousness of the arrested population had gone from a low of 6.7 to a high of 7.34, this would be considered as a 10 percent increase in the level of dangerousness. 7. It should be noted that validation data for the distribution of criminal history and violence history were only available for the prison and parole populations; the distributions of these indicators had to be estimated for other system populations. The details of the estimation procedures used are described in the Technical Appendix. 8. As just noted, this discussion is necessarily speculative. Apart from the fact that the distribution of criminal and violence histories was estimated for all system populations except prison and parole (see the Technical Appendix for details of the estimation procedures), the methodology employed here does not justify the imputation of group trends in a single item of the dangerousness construct to the level of individual offenders. 9. This portion of the discussion is necessarily speculative, as complete data on subgroups were not available on the probation population. These data were estimated, based on a number of assumptions about the composition of the probation population (for example, no offenders with two prior convictions were included in the probation
Notes to Chapter 8
185
population) and on arrest volume. The linkage of the probation estimates to the actual arrest data in this fashion increases confidence in the estimates. Data sources and estimation procedures are detailed further in the Technical Appendix. 10. The careful observer will note that the male series displays a much smoother trend than the female series. One possible explanation for the two sharp spikes in the rates of growth in the female prison population is the increased capacity afforded by the opening of several new women’s prisons. The 1990 jump likely results from the opening of the Northern California Women’s Facility in 1987 (the first new women’s correctional facility in California since the opening of the California Institution for Women in 1952) and the Central California Women’s Prison in 1990; the 1997 increase may reflect the increased availability of prison beds for women created by the 1995 opening of Valley State Prison for Women. The opening of the Northern California Women’s Facility and the Central California Women’s Facility more than doubled state capacity for women prisoners. I am grateful to Elizabeth Dermody Leonard for suggesting this explanation.
NOTES TO CHAPTER 8 1. The details of the simulation procedures are provided in the technical appendixes. 2. As was shown in chapter 7, gender makes no significant contribution to changes in population dangerousness. In all simulated scenarios, the gender ratio in the prison population remained fixed, with women making up approximately six percent of the California prison population. 3. Programming details are provided in the Technical Appendixes. 4. Programming details are provided in the Technical Appendixes. 5. Although geriatric release might also be defensible from a humanitarian standpoint, such justifications are largely absent from the discourse. For example, in the Department of Justice guidelines for receipt of Violent Offender Incarceration/Truth in Sentencing (VOI/TIS) funding, geriatric release is presented from the perspective of social defense. Additionally, the guidelines state that “the Governor may also release prisoners whose medical condition precludes them from posing a threat to the public” (U.S. Department of Justice 1998:3). 6. It should also be mentioned that geriatric release does not necessarily relieve the state of the costs posed by these offenders. Many of these offenders may be, at the time of release, in possession of limited employment skills and/or ability to work or otherwise support themselves, so the implementation of geriatric release may well simply result in a displacement of costs to other state agencies, rather than a net financial savings. 7. This age group was also selected for pragmatic reasons, given that this is the cutoff point for age groups in the simulation modeling code. Programming details of the simulation modeling are provided in the Technical Appendix.
186
Notes to Technical Appendix A
8. The decrease in dangerousness seen in the geriatric release scenario is actually slightly larger than that seen in the simulated Three Strikes full implementation scenario. This is an artifact of the spontaneous generation of the system in each simulation run. 9. Although these features of the Three Strikes law do not appear to contribute to the preservation of the public safety, this does not mean that the law might not be defensible on other grounds—such as deterrence or retribution.
NOTES TO TECHNICAL APPENDIX A 1. Available for demonstration and purchase at www.berkeleymadonna.com. 2. It is important to note that at this level of the research endeavor, the exact reasons for the variability are not important. Put another way, if, for example, black males aged 18–24 have greater rates of pretrial detention overall than do white women aged 25–34, the mechanisms (e.g., racist judges, differences in the ability of offenders to pay cash bail) that bring this reality about are epiphenominal. It should be remembered that the sole purpose of the model presented in chapter 7 is to reproduce and evaluate the historical realities of the composition of the California criminal justice system, and not the testing of any particular theory or hypothesis. 3. This is unfortunate, but not disastrous, as the specific composition of the populations occupying these states are not the primary focus of the model. Luckily, complete data are available on the criminal justice system populations that represent the main focus of this analysis, the prison and parole populations.
Bibliography
Alexander, Jack, 1986. “Classification Objectives and Practices.” Crime and Delinquency 32:3:323–338. Allen, Francis A., 1981. The Decline of the Rehabilitative Ideal: Penal Policy and Social Purpose. New Haven: Yale University Press. Allen, Hilary, 1987. “Rendering Them Harmless: The Professional Portrayal of Women Charged with Serious Violent Crimes,” pp. 81–94 in Pat Carlen and Anne Worrall (eds.), Gender, Crime, and Justice. Milton Keynes: Open University Press. Allswang, John M., 1991. California Initiatives and Referendums, 1912–1990: A Survey and Guide to Research. Los Angeles: Edmund G. “Pat” Brown Institute of Public Affairs, California State University, Los Angeles. Ancel, Marc, 1987. Social Defense: The Future of Penal Reform. Littleton, CO: Fred B. Rothman. Applegate, Brandon K., Francis T. Cullen, and Bonnie S. Fisher, 1997. “Public Support for Correctional Treatment: The Continuing Appeal of the Rehabilitative Ideal.” Prison Journal 77:237–258. Ashman, Allan, 1979. “Criminal Law: Mandatory Sentences.” American Bar Association Journal 65:1564–1565. Assembly Committee on Criminal Justice, 1979. Analysis of Proposition 8: The Criminal Justice Initiative. Sacramento: California Legislative Assembly. ———, 1982. Hearing on Determinate Sentencing. Sacramento: California Legislative Assembly. Auerhahn, Kathleen, 2002. “Selective Incapacitation, Three Strikes, and the Problem of Aging Prison Populations”. Criminology and Public Policy 1:3:353–388. ———, 1999a. “Selective Incapacitation and the Problem of Prediction.” Criminology 37:4:703–733. ———, 1999b. “The Split Labor Market and the Origins of Antidrug Legislation in the United States.” Law and Social Inquiry 24:2:411–440.
187
188
Bibliography
Austin, James, 1983. “Assessing the New Generation of Prison Classification Models.” Crime and Delinquency 29:561–576. ———, 1986. “Evaluating How Well Your Classification System Is Operating: A Practical Approach.” Crime and Delinquency 32:3:302–322. Austin, James, John Clark, Patricia Hardyman, and D. Alan Henry, 1999. “The Impact of ‘Three Strikes and You’re Out.” Punishment and Society 1:2:131–162. Austin, James, and William Pannell, 1985. The Growing Imprisonment of California. San Francisco: National Council on Crime and Delinquency. Avi-Itzhak, Benjamin, and Reuel Shinnar, 1973. “Quantitative Models in Crime Control.” Journal of Criminal Justice 1:185–217. Axelrod, Robert, 1995. “A Model of the Emergence of New Political Actors,” pp. 19–39 in Nigel Gilbert and Rosaria Conte (eds.), Artificial Societies: The Computer Simulation of Social Life. London: UCL Press. Bales, William D., and Linda G. Dees, 1992. “Mandatory Minimum Sentencing in Florida: Past Trends and Future Implications.” Crime and Delinquency 38:3:309–329. Baum, Dan, 1996. Smoke and Mirrors: The War on Drugs and the Politics of Failure. Boston: Little, Brown. Baum, Noah, and Brooke Bedrick, 1994. Trading Books for Bars: The Lopsided Funding Battle between Prisons and Universities. San Francisco: Center on Juvenile and Criminal Justice. Bean, Walton, and James J. Rawls, 1983. California: An Interpretive History. 4th ed. New York: McGraw-Hill. Beccaria, Cesare, 1775 [1983]. An Essay on Crimes and Punishments. Boston: Branden Publishing. Beck, Ulrich, 1992. Risk Society: Towards a New Modernity. Translated by Mark Ritter. London: Sage. Beirne, Piers, 1993. Inventing Criminology: Essays on the Rise of ‘Homo Criminalis.’ Albany: State University of New York Press. Belkin, Jacob, Alfred Blumstein, William Glass, and Michael Lettre, 1972. “JUSSIM, An Interactive Computer Program and its Uses in Criminal Justice Planning,” pp. 467–477 in Proceeding, Project SEARCH. Sacramento, CA: Committee on Security and Privacy. Benekos, Peter J., and Alida V. Merlo, 1995. “Three Strikes and You’re Out! The Political Sentencing Game.” Federal Probation 59:1:3–9. Berk, Richard A., Harold Brackman, and Selma Lesser, 1977. A Measure of Justice: An Empirical Study of Changes in the California Penal Code, 1955–1971. New York: Academic.
Bibliography
189
Bertram, Eva, Morris Blachman, Kenneth Sharpe, and Peter Andreas, 1996. Drug War Politics: The Price of Denial. Berkeley: University of California Press. Bjørkly, Stål, 1993. “Scale for the Prediction of Aggression and Dangerousness in Psychotic Patients, an Introduction.” Psychological Reports 73:1363–1377. Black, Donald, 1970. “Production of Crime Rates.” American Sociological Review 35:733–748. Blackmore, John, and Jane Welsh, 1983. “Selective Incapacitation: Sentencing according to Risk.” Crime and Delinquency 29:504–529. Bloom, Barbara, Meda Chesney–Lind, and Barbara Owen, 1994. Women in California Prisons: Hidden Victims of the War on Drugs. San Francisco: Center on Juvenile and Criminal Justice. Bookspan, Shelley, 1991. A Germ of Goodness: The California State Prison System, 1851–1944. Lincoln: University of Nebraska Press. Blumstein, Alfred, 1982. “On the Racial Disproportionality of United States’ Prison Populations.” Journal of Criminal Law and Criminology 73:3:1259–1281. ———, 1983. “Prisons: Population, Capacity, and Alternatives,” pp. 229–250 in James Q. Wilson (ed.), Crime and Public Policy. San Francisco: Institute for Contemporary Studies Press. Blumstein, Alfred, and Gary C. Koch, 1977. “A Prolegomenon for a Macro Model for Criminal Justice Planning: JUSSIM III.” Criminal Justice Planning. Blumstein, Alfred, Jacqueline Cohen, and Daniel Nagin (eds.), 1978. Deterrence and Incapacitation: Estimating the Effects of Criminal Sanctions on Crime Rates. Washington DC: National Academy of Sciences. Blumstein, Alfred, Jacqueline Cohen, Jeffrey A. Roth, and Christy A. Visher (eds.), 1986. Criminal Careers and “Career Criminals.” Vol. 2. Washington, DC: National Academy Press. Blumstein, Alfred, and Elizabeth Graddy 1982. “Prevalence and Recidivism in Index Arrests: A Feedback Model.” Law and Society Review 16:2:265–290. Blumstein, Alfred, and Richard Larson, 1969. “Models of a Total Criminal Justice System.” Operations Research 17:199–232. Boswell, Terry E., 1986. “A Split Labor Market Analysis of Discrimination against Chinese Immigrants, 1850–1882.” American Sociological Review 51:352–371. Braithwaite, John, 1989. “The State of Criminology: Theoretical Decay or Rennaissance?” Australia and New Zealand Journal of Criminology September:129–135. Brewer, David, Gerald E. Beckett, and Norman Holt, 1981. “Determinate Sentencing in California: The First Year’s Experience.” Journal of Research in Crime and Delinquency 18:2:200–231.
190
Bibliography
Bridges, George S., and Robert D. Crutchfield, 1988. “Law, Social Standing, and Racial Disparities in Imprisonment.” Social Forces 66:3:699–724. Bridges, George S., and Sara Steen, 1998. “Racial Disparities in Official Assessments of Juvenile Offenders: Attributional Stereotypes as Mediatiing Mechanisms.” American Sociological Review 63:4:554–570. Briggs, John, Christopher Harrison, Angus McInnes, and David Vincent, 1996. Crime and Punishment in England: An Introductory History. New York: St. Martin’s. Brown, M. Craig, and Barbara D. Warner, 1992. “Immigrants, Urban Politics, and Policing in 1900.” American Sociological Review 57:293–305. Buchanan, Robert A., Karen L. Whitlow, and James Austin, 1986. “National Evaluation of Objective Prison Classification Systems: The Current State of the Art.” Crime and Delinquency 32:3:272–290. Bureau of Justice Assistance, 1998. 1996 National Survey of State Sentencing Structures. Bureau of Justice Statistics, 2001. “Nation’s State Prison Population Falls in Second Half of 2000.” [press release]. U.S. Department of Justice. Available online at http://www. ojp.usdoj.gov/bjs/pub/press/p00pr.htm. ———, 1999. Correctional Populations in the United States, 1996. U.S. Department of Justice. Byrne, David, 1997. “Simulation—A Way Forward?” Sociological Research Online 2:2. Available online at http://www.socresonline.org.uk/socresonline/2/2/4.html. California Board of Corrections, Jail Planning and Construction Division, 1987. State of the Jails in California: Report #4. Sacramento: Board of Correction. California Department of Corrections, 1995. 50 Years: Public Safety, Public Service. Sacramento. ———, 2000. CDC Facilities. Available online at http://www.cdc.state.ca.us/facility/ facil.html. ———, 2002. CDC Facts, First Quarter 2002. Available online at http://www.cdc. state.ca.us/factsht.htm. California Department of Justice, 1997. California Criminal Justice Profile 1996. Sacramento: Criminal Justice Statistics Center. ———, 2002. California Criminal Justice Profile 2000. Available online at http://caag.state. ca.us/cjsc/datatabs.htm. ———, 1981. California Criminal Justice Profile 1980. Sacramento: Criminal Justice Statistics Center. California District Attorneys Association, 1979. The Uniform Determinate Sentencing Act of California. 2d ed. Sacramento: California District Attorneys Association.
Bibliography
191
Canela-Cacho, José A., Alfred Blumstein, and Jacqueline Cohen, 1997. “Relationship between the Offending Frequency () of Imprisoned and Free Offenders.” Criminology 35:1:13–175. Cantor, Nathaniel F., 1932. Crime, Criminals, and Criminal Justice. New York: H. Holt and Company. Carey, William T., 1979. “Determinate Sentencing in California and Illinois: Its Effect on Sentence Disparity and Prisoner Rehabilitation.” Washington University Law Quarterly 2:551–569. Cary, Seth C., 1889. “Prison Reform: The Indeterminate Sentence.” Alpha (Boston University) 8:1:1–5. Carlson, James M., and Tricia Williams, 1993. “Perspectives on the Seriousness of Crimes.” Social Science Research 22:190–207. Casper, Jonathan D., David Brereton, and David Neal, 1982. The Implementation of the California Determinate Sentencing Law. Washington, DC: National Institute of Justice. ———, 1983. “The California Determinate Sentence Law.” Criminal Law Bulletin 19:405–433. Cassidy, R. Gordon, 1985. “Modelling a Criminal Justice System,” pp. 193–207 in David P. Farrington and Roger Tarling, 1985 (eds.).,Prediction in Criminology. Albany: State University of New York Press. Cassidy, R. Gordon, Michael H. Peters, and Ronald E. Turner, 1981. “Interrelation of Criminal Justice System Resources and Workload.” Infor 19:2:91–112. Cassidy, R. Gordon, and Ronald E. Turner, 1978. “Criminal Justice System Behavior.” Behavioral Science 23:99–108. Cassou, April Kestell, and Brian Taugher, 1978. “Determinate Sentencing in California: The New Numbers Game.” Pacific Law Journal 9:5–106. Caulkins, Jonathan P., C. Peter Rydell, William L. Schwabe, and James Chiesa, 1997. Mandatory Minimum Drug Sentences: Throwing away the Key or the Taxpayer’s Money? Santa Monica, CA: Rand. Chaiken, Marcia R., and Jan Chaiken, 1984. “Offender Types and Public Policy.” Crime and Delinquency 30:195–226. ———, 1990. Redefining the Career Criminal: Priority Prosecution of High-rate Dangerous Offenders. Washington, DC: National Institiute of Justice. Chaiken Jan M., and Marcia R. Chaiken, 1982. Varieties of Criminal Behavior. Santa Monica: CA: Rand. Chapin, Bradley, 1983. Criminal Justice in Colonial America, 1606–1660. Athens: University of Georgia Press. Chesney-Lind, Meda, and Randall G. Shelden, 1992. Girls, Delinquency, and Juvenile Justice. Pacific Grove, CA: Brooks/Cole.
192
Bibliography
Childress, Michael, 1994a. A System Description of the Marijuana Trade. Santa Monica, CA: RAND. ———, 1994b. A System Description of the Heroin Trade. Santa Monica, CA: RAND. Christie, Nils, 1994. Crime Control as Industry. London: Routledge. Clark, John, James Austin, and D. Alan Henry, 1997. Three Strikes and You’re Out: A Review of State Legislation. National Institute of Justice. Available online at http://www.ncjrs.org. Clear, Todd R., 1994. Harm in American Penology: Offenders, Victims, and Their Communities. Albany: State University of New York Press. ———, 1996. “Science and the Punishment/Control Movement.” Social Pathology 2:1–22. Clear, Todd R., and Donald M. Barry, 1983. “Some Conceptual Issues in Incapacitating Offenders.” Crime and Delinquency 29:529–545. Clinard, Marshall B., 1990. Corporate Corruption: The Abuse of Power. New York: Praeger. Clinard, Marshall B., and Richard Quinney, 1967. Criminal Behavior Systems: A Typology. New York: Holt. Cohen, Jacqueline, 1978. “The Incapacitative Effects of Imprisonment: A Critical Review of the Literature,” pp. 187–243 in Alfred Blumstein, Jacqueline Cohen, and Daniel Nagin, (eds.), 1978. Deterrence and Incapacitation: Estimating the Effects of Criminal Sanctions on Crime Rates. Washington, DC: National Academy of Sciences. ———, 1983. “Incapacitation as a Strategy for Crime Control: Possibilities and Pitfalls,” in Michael Tonry and Norval Morris (eds.), Crime and Justice: An Annual Review of Research. Vol. 5. Chicago: University of Chicago Press. ———, 1986. Incapacitating Criminals: Recent Research Findings. National Institute of Justice Research in Brief. Washington, DC: National Institute of Justice. Cohen, Jacqueline, Kenneth Fields, Michel Lettre, Richard Stafford, and Claire Walker, 1973. “Implementation of the JUSSIM Model in a Criminal Justice Planning Agency.” Journal of Research in Crime and Delinquency 10:2:117–131. Cohen, Lawrence E., and Marcus Felson, 1979. “Social Change and Crime Rate Trends: A Routine Activities Approach.” American Sociological Review 44:4:588–608. Cohen, Stanley, 1988. Against Criminology. New Brunswick and Oxford: Transaction Books. Cohn, Ellen G., and David P. Farrington, 1994. “Who Are the Most–Cited Scholars in Major American Criminology and Criminal Justice Journals?” Journal of Criminal Justice 22:6:517–534. ———, 1998. “Changes in the Most-Cited Scholars in Major American Criminology and Criminal Justice Journals between 1986–1990 and 1991–1995.” Journal of Criminal Justice 26:2:99–116.
Bibliography
193
Congressional Information Service, 1984, 1986. Abstracts of Congressional Publications. Connolly, Kathleen, Lea McDermid, Vincent Schiraldi, and Dan Macallair, 1996. From Classrooms to Cell Blocks: How Prison Building Affects Higher Education and African American Enrollment. San Francisco: Center on Juvenile and Criminal Justice. Conrad, John P., 1982. “The Quandary of Dangerousness.” British Journal of Criminology 22:3:255–267. Conrad, John P., and Simon Dinitz, 1977. In Fear of Each Other: Studies of Dangerousness in America. Lexington, MA: Lexington Books. Conte, Rosaria, and Nigel Gilbert, 1995. “Introduction: Computer Simulation for Social Theory,” pp. 1–15 in Nigel Gilbert and Rosaria Conte (eds.), Artificial Societies: The Computer Simulation of Social Life. London: UCL Press. Copas, John, 1983. “Some Statistical Questions in the Prediction of Dangerous Offending,” in John W. Hinton (ed.), Dangerousness: Problems of Assessment and Prediction. London: Allen & Unwin. ———, 1985. “Prediction Equations, Statistical Analysis, and Shrinkage,” pp. 232–255 in David P. Farrington and Roger Tarling, 1985 (eds.). Prediction in Criminology. Albany: State University of New York Press. Corrections Digest, 1980. “Determinate Sentencing Law SB 42 Overcrowds California Prisons.” Volume 11, January 4, 1980. Annandale, VA: Washington Crime News Services. Costello, Anthony, Rick Garnett, and Vincent Schiraldi, 1991. Parole Violators in California: A Waste of Money, A Waste of Time. San Francisco: Center on Juvenile and Criminal Justice. Coulson, Grant, Giorgio Ilacqua, Verna Nutbrown, Diana Guilekas, and Francis Cudjoe, 1996. “Predictive Utility of the LSI for Incarcerated Female Offenders.” Criminal Justice and Behavior 23:3:427–439. Crawford, Charles, Ted Chiricos, and Gary Kleck, 1998. “Race, Racial Threat, and Sentencing of Habitual Offenders.” Criminology 36:3:481–511. Crutchfield, J. P., 1992. “Knowledge and Meaning: Chaos and Complexity,” pp. 66–101 in Lui Lam and Vladimir Naroditsky (eds.), Modeling Complex Phenomena: Proceedings of the Third Woodward Conference. New York: Springer–Verlag. Cullen, Francis T., and Karen E. Gilbert, 1982. Reaffirming Rehabilitation. Cincinnati, OH: Anderson. Currie, Elliott, 1992. “Retreatism, Minimalism, Realism: Three Styles of Reasoning on Crime and Drugs in the United States,” pp. 88–97 in John Lowman and Brian D. MacLean (eds.), Realist Criminology: Crime Control and Policing in the 1990s. Toronto: University of Toronto Press. Dahrendorf, Ralf , 1958. “Out of Utopia: Toward a Reorientation of Sociological Analysis.” American Journal of Sociology 64:115–127.
194
Bibliography
———, 1959. Class and Class Conflict in Industrial Society. Stanford: Stanford University Press. Davies, J., 1980. “The London Garroting Panic of 1862,” in V.A.C. Gattrell, Bruce Lenman, and Geoffrey Parker (eds.), Crime and Law: The Social History of Crime in Western Europe since 1500. London: European Publications Ltd. Davis, Christopher, Richard Estes, and Vincent Schiraldi, 1996. Three Strikes: The New Apartheid. San Francisco: Center on Juvenile and Criminal Justice. Davis, Mike, 1995. “Hell Factories in the Field: A Prison–Industrial Complex.” Nation 260:7:229–234. Dershowitz, Alan M., 1973. “Preventive Confinement: A Suggested Framework for Constitutional Analysis.” Texas Law Review 51:1277–1324. ———, 1976. “Background Paper,” pp. 67–130 in Fair and Certain Punishment: Report of the Twentieth Century Fund, Task Force on Criminal Sentencing. New York: McGraw–Hill. Deschner, Jeanne P., Michael D. Plain, Gerald K. Terhune, and Celia Williamson, 1981. “Development of the DDI: A Seriousness Scale for Delinquency.” Evaluation Review 5:6:788–809. DeWolf, L. Harold, 1975. Crime and Justice in America: A Paradox of Conscience. New York: Harper. Dombey–Moore, Bonnie, Susan Resetar, and Michael Childress, 1994. A System Description of the Cocaine Trade. Santa Monica, CA: RAND. Dorny, C. Nelson, 1993. Understanding Dynamic Systems: Approaches to Modeling, Analysis, and Design. Englewood Cliffs, NJ: Prentice–Hall. Douglas, Mary, 1992. Risk and Blame: Essays in Cultural Theory. London: Routledge. Dubber, Markus Dirk, 1990. “The Unprincipled Punishment of Repeat Offenders: A Critique of California’s Habitual Criminal Statute.” Stanford Law Review 43:193–240. Durkheim, Emile, 1982. The Rules of Sociological Method. Translated by W. D. Halls. New York: Free Press. ———, 1984. The Division of Labor in Society. Translated by W. D. Halls. New York: Free Press. Elias, Norbert, 1982 [1939]. Power and Civility. Vol. 2, The Civilizing Process. Translated by Edmund Jephcott with notes and revisions by the author. New York: Pantheon. ———, 1988. “Violence and Civilization: The State Monopoly of Physical Violence and its Infringement,” pp 177–198 in John Keane (ed.), Civil Society and the State: New European Perspectives. London: Verso.
Bibliography
195
Ellingwood, Herbert, 1985. Law Enforcement and Criminal Justice in California, 1966–1976: Interviews Conducted by Gabrielle Morris and Sarah Sharp. Berkeley: Regional Oral History Office, Bancroft Library, University of California. Epstein, Joshua M., and Robert Axtell, 1996. Growing Artificial Societies: Social Science from the Bottom Up. Washington, DC: Brookings Institution Press. Erikson, Kai T., 1966. Wayward Puritans: A Study in the Sociology of Deviance. Boston: Allyn. Eve, Raymond A., Sara Horsfall, and Mary E. Lee (eds.), 1997. Chaos, Complexity, and Sociology: Myths, Models, and Theories. Thousand Oaks, CA: Sage. Farrington, David P., 1979. “Longitudinal Research on Crime and Delinquency, pp. 289–348 in Norval Morris and Michael Tonry (eds.), Crime and Justice: An Annual Review of Research. Vol. 1. Chicago and London: University of Chicago Press. ———, 1986. “Age and Crime,” in Norval Morris and Michael Tonry (eds.), Crime and Justice: An Annual Review of Research. Vol. 7. Chicago and London: University of Chicago Press. Farrington, David P., and Roger Tarling, 1985a. ”Criminological Prediction: An Introduction,” in David P. Farrington and Roger Tarling (eds.), Prediction in Criminology. Albany: State University of New York Press. ———, 1985b (eds.), Prediction in Criminology. Albany: State University of New York Press. Feeley, Malcolm M., and Jonathan Simon, 1992. “The New Penology: Notes on the Emerging Strategy of Corrections and Its Implications.” Criminology 30:4:449–474. Fernandez, Kenneth E., and Max Neiman, 1998. “California’s Inmate Classification System: Predicting Inmate Misconduct.” Prison Journal 78:4:406–422. Ferraro, Kenneth F., 1995. Fear of Crime: Interpreting Victimizaton Risk. Albany: State University of New York Press. Figlio, Robert M., 1996. “The Continuation of Juvenile Delinquency into Adult Criminality: Some Findings from Interviews of Members of the 1958 Birth Cohort.” Pepperdine Law Review 23:4:843–852. Flanagan, Timothy J., 1996. “Reform or Punish: Americans’ Views of the Correctional System,” pp. 75–92 in Timothy J. Flanagan and Dennis R. Longmire (eds.), Americans View Crime and Justice: A National Public Opinion Survey. Thousand Oaks, CA: Sage. Flanagan, Timothy J., and S. R. Caulfield, 1984. “Public Opinion and Prison Policy: A Review.” Prison Journal 64:39–59. Floud, Jean, and Warren Young, 1981. Dangerousness and Criminal Justice. Totowa, NJ: Barnes and Noble Books. Foote, Caleb, 1993. The Prison Population Explosion: California’s Rogue Elephant. San Francisco: Center on Juvenile and Criminal Justice.
196
Bibliography
Forrester, Jay W., 1961. Industrial Dynamics. Cambridge: Massachusetts Institute of Technology Press. ———, 1968. Principles of Systems. Cambridge: Massachusetts Institute of Technology Press. ———, 1969a. Urban Dynamics. Cambridge: Massachusetts Institute of Technology Press. ———, 1969b. “A Deeper Knowledge of Social Systems.” Technology Review 71:6:21–30. ———, 1973. World Dynamics. Cambridge: Massachusetts Institute of Technology Press. Foster, Michele, 1996. “As California Goes, So Goes the Nation.” Journal of Negro Education 65:2:105–110. Foucault, Michel, 1977. Discipline and Punish: The Birth of the Prison. Translated by Alan Sheridan. New York: Vintage. ———, 1988. “The Dangerous Individual,” pp. 125–151 in Lawrence D. Kritzman (ed.), Politics, Philosophy, Culture: Interviews and Other Writings 1977–1984. New York: Routledge. ———, 1988b. “Confinement, Psychiatry, Prison,” pp. 178–210 in Lawrence D. Kritzman (ed.), Politics, Philosophy, Culture: Interviews and Other Writings 1977–1984. New York: Routledge. Frankel, Marvin E., 1973. Criminal Sentences: Law Without Order. New York: Hill and Wang. Friedman, Lawrence M., 1993. Crime and Punishment in American History. New York: Basic. Friedrich, Barbara, 1999. California Ballot Propositions 1944–present. Compiled and Edited by Barbara Friedrich, Head of Reference, Heafey Law Library, Santa Clara University School of Law. Available online at http://www.scu.edu/law/pubs/Props.htm. Garland, David, 1985. Punishment and Welfare: A History of Penal Strategies. Aldershot, England: Gower. ———, 1990. Punishment and Modern Society: A Study in Social Theory. Chicago: University of Chicago Press. Geis, Gilbert, 1968. White-Collar Criminal: The Offender in Business and Professions. New York: Atherton Press. Geis, Gilbert, and Ivan Bunn, 1997. A Trial of Witches: A Seventeenth Century Witchcraft Prosecution. London and New York: Routledge. Gendreau, Paul, Patrick G. Madden, and Mary Leipciger, 1980. “Predicting Recidivism with Social History Information and a Comparison of Their Predictive Power with Psychometric Variables.” Canadian Journal of Criminology 22:3:328–336.
Bibliography
197
Gerber, Jurg, and Simone Engelhardt–Greer, 1996. “Just and Painful: Attitudes Toward Sentencing Criminals,” pp. 62–74 in Timothy J. Flanagan and Dennis R. Longmire (eds.), Americans View Crime and Justice: A National Public Opinion Survey. Thousand Oaks, CA: Sage. Gibbs, Jack P., 1975. Crime, Punishment, and Deterrence. New York: Elsevier. Gilbert, Nigel, 1995. “Emergence in Social Simulation,” pp. 144–156 in Nigel Gilbert and Rosaria Conte (eds.), Artificial Societies: The Computer Simulation of Social Life. London: UCL Press. Gilbert, Nigel, and Rosaria Conte (eds.), 1995. Artificial Societies: The Computer Simulation of Social Life. London: UCL Press. Gillard, Darrell K., and Allen J. Beck, 1998. Prison and Jail Inmates at Midyear 1997. Washington DC: Bureau of Justice Statistics, U.S. Department of Justice. Gilmore, Ruth Wilson, 1998. “Globalisation and U.S. Prison Growth: From Military Keynesiansim to Post–Keynesian Militarism.” Race and Class 171. Glaser, Daniel, 1995. Preparing Convicts for Law–Abiding Lives: The Pioneering Penology of Richard A. McGee. Albany: State University of New York Press. Glick, Henry R., 1979. “Mandatory Sentencing: The Politics of the New Criminal Justice.” Federal Probation 43:1:3–9. Gordon, Diana R., 1994. The Return of the Dangerous Classes: Drug Prohibition and Policy Politics. New York: Norton. Gordon, Robert A., 1977. “A Critique of the Evaluation of the Patuxent Institution, with Particular Attention to the Issues of Dangerousness and Recidivism.” Bulletin of the American Academy of Psychiatry and the Law 5:210–255. Gorton, Joe, and John L. Boies, 1999. “Sentencing Guidelines and Racial Disparity Across Time: Pennsylvania Prison Sentences in 1977, 1983, 1992, and 1993.” Social Science Quarterly 80:1:37. Gottfredson, Don M., and Michael Tonry (eds.), 1987. Prediction and Classification: Criminal Justice Decision Making. Crime and Justice: A Review of Research. Vol. 9. Chicago: University of Chicago Press. Gottfredson, Michael R. and Travis Hirschi, 1990. A General Theory of Crime. Stanford: Stanford University Press. Gottfredson, Stephen D., 1987. “Prediction: An Overview of Selected Methodological Issues,” pp. 21–51 in Don M. Gottfredson and Michael Tonry (eds.), Prediction and Classification: Criminal Justice Decision Making. Chicago: University of Chicago Press. Gottfredson, Stephen D., and Don M. Gottfredson, 1985a. “Selective Incapacitation?” Annals of the American Academy of Political and Social Sciences 478: March:135–149.
198
Bibliography
———, 1985b. “Screening for Risk Among Parolees: Policy, Practice, and Method,” pp. 54–77 in David P. Farrington and Roger Tarling, 1985 (eds.), Prediction in Criminology. Albany: State University of New York Press. ———, 1992. Incapacitation Strategies and the Criminal Career. Sacramento: California Department of Justice. ———, 1994. “Behavioral Prediction and the Problem of Incapacitation.” Criminology 32:3:441–474. Greenberg, David F., 1992. “Comparing Criminal Career Models.” Criminology 30:1:141–147. Greenberg, David, 1975. “The Incapacitative Effect of Imprisonment: Some Estimates.” Law and Society Review 9:541–80. Greenfield, Lawrence A., and Stephanie Minor-Harper, 1991. Women in Prison. Washington, DC: Bureau of Justice Statistics, U.S. Department of Justice. Greenwood, Peter W., 1983. “Controlling the Crime Rate Through Imprisonment,” pp. 251–269 in James Q. Wilson (ed.), Crime and Public Policy. San Francisco: Institute for Contemporary Studies Press. Greenwood, Peter W., with Allan Abrahamse, 1982. Selective Incapacitation. Santa Monica, CA: RAND. Greenwood, Peter W., Allan Abrahamse, Susan Everingham, Elsa Y. Chen, Nancy Merritt, and James Chiesa, 1998. Measuring the Impact of Three Strikes and You’re Out at the National, State, and County Level. Santa Monica, CA: Rand. Greenwood, Peter W., C. Peter Rydell, Allan F. Abrahamse, Jonathan P. Caulkins, James Chiesa, Karyn E. Model, and Stephen P. Klein, 1994. Three Strikes and You’re Out: Estimated Benefits and Costs of California’s New Mandatory-Sentencing Law. Santa Monica, CA: RAND. Griset, Pamala L., 1996. “Determinate Sentencing and Administrative Discretion Over Time Served in Prison: A Case Study of Florida.” Crime and Delinquency 42:1:127–143. Griswold, David B., and Michael D. Wiatrowksi, 1983. “The Emergence of Determinate Sentencing.” Federal Probation 47:2:28–35. Guerin-Gonzales, Camille, 1994. Mexican Workers and American Dreams: Immigration, Repatriation, and California Farm Labor, 1900–1939. New Brunswick, NJ: Rutgers University Press. Gusfield, Joseph R., 1963. Symbolic Crusade: Status Politics and the American Temperance Movement. Urbana: University of Illinois Press. Hakeem, Michael, 1948. “The Validity of the Burgess Method of Parole Prediction.” American Journal of Sociology 53:5:376–386. Hale, Chris, 1996. “Fear of Crime: A Review of the Literature.” International Review of Victimology 4:2:79–150.
Bibliography
199
Hamparian, Donna Martin, Richard Schuster, Simon Dinitz, and John P. Conrad, 1978. The Violent Few: A Study of Dangerous Juvenile Offenders. Lexington, MA: Lexington Books. Hanneman, Robert A., 1988. Computer-Assisted Theory Building: Modelling Dynamic Social Systems. Beverly Hills: Sage. ———, 1995. “Simulation Modeling and Theoretical Analysis in Sociology.” Sociological Perspectives 38:4:457–462. Hanneman, Robert A., Randall Collins, and Gabrielle Mordt, 1995. “Discovering Theory Dynamics by Computer Simulation: Experiments on State Legitimacy and Imperialist Capitalism,” pp. 1–46 in Peter Marsden (ed.), Sociological Methodology 1995. Washington, DC: American Sociological Association. Hanneman, Robert A., and Steven Patrick, 1997. “On the Uses of Computer-Assisted Simulation Modeling in the Social Sciences.” Sociological Research Online 2:2. Available online at http://www.socresonline.org.uk/socresonline/2/2/5.html. Harring, Sidney L., 1977. “Class Conflict and the Suppression of Tramps in Buffalo, 1892–1894.” Law and Society Review 11:873–911. Hart, H.L.A., 1968. Punishment and Responsibility: Essays in the Philosophy of Law. Oxford: Clarendon. Hatch, Orrin G., 1993. “The Role of Congress in Sentencing: The United States Sentencing Commission, Mandatory Minimum Sentences, and the Search for a Certain and Effective Sentencing System.” Wake Forest Law Review 28:185–198. Hepburn, John R., and Lynne Goodstein, 1986. “Organizational Imperatives and Sentencing Reform Implementation: The Impact of Prison Practices and Priorities on the Attainement of the Objective of Determinate Sentencing.” Crime and Delinquency 32:3:339–365. Hewitt, John D., and Todd R. Clear, 1983. The Impact of Sentencing Reform: From Indeterminate to Determinate Sentencing. Lanham, MD: University Press of America. Hibbert, Christopher, 1963. The Roots of Evil: A Social History of Crime and Punishment. New York: Minerva Press. High Performance Systems, 1997. STELLA: An Introduction to Systems Thinking. High Performance Systems. Hindelang, Michael J., 1978. “Race and Involvement in Common Law Personal Crimes.” American Sociological Review 43:1:93–109. Hoffman, Peter B., 1982. “Females, Recidivism, and Salient Factor Score: A Research Note.” Criminal Justice and Behavior 9:1:121–125. ———, 1994. “Twenty Years of Operational Use of a Risk Prediction Instrument: The United States Parole Commission’s Salient Factor Score.” Journal of Criminal Justice 22:6:477–494.
200
Bibliography
Hoffman, Peter B., and James L. Beck, 1985. “Recidivism among Released Federal Prisoners: Salient Factor Score and Five–Year Follow–Up.” Criminal Justice and Behavior 12:4:501–507. Holmes, Robert E., 1992. Sentencing in California: Where Are We Today? Where Are We Going? Sacramento: California Research Bureau. Howard, Dick, 1978. Determinate Sentencing in California. Lexington, KY: Council of State Governments. Howard, George W., and John J. McHugh, 1978. “Fixed Sentences Emerge as New Trend in Corrections.” American Journal of Correction 40:4:30–34. Howe, Edmund S., 1994. “Judged Person Dangerousness as Weighted Averaging.” Journal of Applied Social Psychology 24:14:1270–1290. Ignatieff, Michael, 1978. A Just Measure of Pain: The Penitentiary in the Industrial Revolution, 1750–1850. New York: Pantheon. ———, 1981. “State, Civil Society, and Total Institutions: A Critique of Recent Social Histories of Punishment,” pp. 153–191 in Michael Tonry and Norval Morris (eds.), Crime and Justice: An Annual Review of Research. Chicago: University of Chicago Press. Innes, Christopher A., 1993. “Recent Public Opinion in the United States toward Punishment and Corrections.” Prison Journal 73:2:220–236. Irwin, John, 1985. The Jail: Managing the Underclass in American Society. Berkeley: University of California Press. Irwin, John, and James Austin, 1994. It’s about Time: America’s Imprisonment Binge. Belmont, CA: Wadsworth. Jackson, George, 1970. Soledad Brother: The Prison Letters of George Jackson. New York: Coward-McCann. Jackson, Margaret Windsor, 1988. “Lay and Professional Perceptions of Dangerousness and Other Forensic Issues.” Canadian Journal of Criminology 215–229. Jackson, Pamela Irving, 1986. “Black Visibility, City Size, and Social Control.” Sociological Quarterly 27:2:185–203. Jacobsen, Chanoch, and Richard Bronson, 1985. Simulating Violators. Arlington, VA: Operations Research Society of America. Jones, Bill, 1998. A History of the California Initiative Process. Sacramento: Secretary of State’s Office. Jones, David A., 1986. History of Criminology: A Philosophical Perspective. Westport, CT: Greenwood Press. Kadish, Sanford H., 1978. “Introduction,” Determinate Sentencing: Reform or Regression? Proceedings of the Special Conference on Determinate Sentencing, Boalt Hall School of Law, Univeristy of California, Berkeley. Washington, DC: Law Enforcement Assistance Administration, U.S. Department of Justice.
Bibliography
201
Kane, Thomas R., 1986. “The Validity of Prison Classification: An Introduction to Practical Considerations and Research Issues.” Crime and Delinquency 32:3:367–390. Kannensohn, Michael, 1980. A National Survey of Parole–Related Legislation Enacted During the 1979 Legislative Session. Washington, DC: Department of Justice. Kempinen, Cynthia, 1997. “Criminal History of Defendants and Sentencing Guidelines: A Comparison of Three Strikes Legislation and Sentencing Guidelines on Identifying the Dangerous Offender.” Paper presented at the annual meeting of the American Society of Criminology, San Diego, California. Kittrie, Nicholas N., 1971. The Right to Be Different: Deviance and Enforced Therapy. Baltimore: Johns Hopkins University Press. Kovandzic, Tomislav V., 2001. “The Impact of Florida’s Habitual Offender Law on Crime.” Criminology 39:1:179–203. Kozol, H. L., R. J. Boucher, and R. F. Garofalo, 1972. “Diagnosis and Treatment of Dangeorusness.” Crime and Delinquency 18:4:372–392. Kuhn, Thomas S., 1996. The Structure of Scientific Revolutions. 3rd ed. Chicago: University of Chicago Press. Kuntz, William Francis (II), 1988. Criminal Sentencing in Three Nineteenth-Century Cities: Social History of Punishment in New York, Boston, and Philadelphia 1830–1880. New York: Garland Publishing. Kuzmics, Helmut, 1988. “The Civilizing Process,” pp 149–176 in John Keane (ed.), Civil Society and the State: New European Perspectives. London: Verso. Lamott, Kenneth, 1961. Chronicles of San Quentin: Biography of a Prison. New York: McKay. Langan, Patrick A., 1985. “Racism on Trial: New Evidence to Explain the Racial Composition of Prisons in the United States.” Journal of Criminal Law and Criminology 76:3:666–683. Langan, Patrick A., John V. Fundis, Lawrence Greenfield, and Victoria W. Schneider, 1988. Historical Statistics on Prisoners in State and Federal Institutions, Yearend 1925–86. U.S. Department of Justice: Bureau of Justice Statistics. Langbein, John H., 1998. “The Historical Origins of the Sanction of Imprisonment for Serious Crime,” in Timothy J. Flanagan, James W. Marquart, and Kenneth G. Adams (eds.), Incarcerating Criminals: Prisons and Jails in Social and Organizational Context. New York: Oxford University Press. Lea, John, 1992. “The Analysis of Crime,” pp. 69–94 in Jock Young and Roger Matthews (eds.), Rethinking Criminology: The Realist Debate. London: Sage. Lea, John, and Jock Young, 1984. What Is to Be Done about Law and Order? Harmondsworth, England: Penguin.
202
Bibliography
Legislative Analyst’s Office, 1995. The “Three Strikes and You’re Out” Law: A Preliminary Assessment. Available online at http://www.lao.ca.gov/sc010695.html. ———, 1999. The “Three Strikes and You’re Out” Law’s Impact on State Prisons: An Update. Available online at http://www.lao.ca.gov. ———, 1997. Addressing the State’s Long-Term Inmate Population Growth. Available online at http://www.lao.ca.gov/acrobat_files/pb052097_addressing_inmate_pop.pdf. Leonard, Elizabeth Dermody, 2002. Convicted Survivors: The Imprisonment of Battered Women Who Kill. Albany: State University of New York Press. Lewis, J. L., 1863. Transportation and Penal Servitude. London: W. Clowes and Sons. Lipton, D., Martinson, R., and Wilks, J., 1975. The Effectiveness of Correctional Treatment: A Survey of Treatment Evaluation Studies. New York: Praeger. Liska, Allen E., and Jiang Yu, 1992. “Specifying and Testing the Threat Hypothesis: Police Use of Deadly Force,” in Allen E. Liska (ed.), Social Threat and Social Control. Albany: State University of New York Press. Litvan, Laura M., 1998. “Can ‘Three Strikes’ Reduce Crime?” Investors Business Daily November 30, 1998; secs. A1 and A26. Lobingier, Charles Sumner, 1987 [1923]. The Evolution of the Roman Law from Before the Twelve Tables to the Corpus Juris. Littelton, CO: Fred B. Rothman & Company. Lotke, Eric, 1996. “The Prison–Industrial Complex.” Multinational Monitor 17:11:18–31. Lowman, John, 1992. “Rediscovering Crime,” pp. 141–160 in Jock Young and Roger Matthews (eds.), Rethinking Criminology: The Realist Debate. London: Sage. Lowrance, William W., 1976. Of Acceptable Risk: Science and the Determination of Safety. Los Altos, CA: William Kaufmann. Maguire, Kathleen, and Ann L. Pastore (eds.), 1995. Sourcebook of Criminal Justice Statisitics 1994. U.S. Department of Justice, Bureau of Justice Statistics. Washington, DC: USGPO. ———, 1997. Sourcebook of Criminal Justice Statistics. Available online at http://www.albany.edu/sourcebook. ———, 2001. Sourcebook of Criminal Justice Statistics. Available online at http://www.albany.edu/sourcebook. Maguire, K., A. L. Pastore, and T. J. Flanagan (eds.), 1993. Sourcebook of Criminal Justice Statistics 1992. U.S. Department of Justice, Bureau of Justice Statistics. Washington, DC: USGPO. Manski, Charles F., 1978. “Prospects for Inference on Deterrence through Empirical Analysis of Individual Criminal Behavior,” pp. 400–424 in Alfred Blumstein, Jacqueline Cohen, and Daniel Nagin (eds.), 1978. Deterrence and Incapacitation: Estimating the Effects of Criminal Sanctions on Crime Rates. Washington DC: National Academy of Sciences.
Bibliography
203
Marquart, James M., Sheldon Ekland-Olson, and Jonathan R. Sorenson, 1989. “Gazing Into the Crystal Ball: Can Jurors Accurately Predict Dangerousness in Capital Cases?” Law and Society Review 23:3:449–468. Martinson, Robert, 1974. “What Works? Questions and Answers about Prison Reform.” Public Interest 35:22–54. ———, 1979. “New Findings, New Views: A Note of Caution Regarding Sentencing Reform.” Hofstra Law Review 7:2:243–258. Mathiesen, Thomas, 1990. Prison on Trial: A Critical Assessment. London: Sage. Matthews, Roger, and Jock Young, 1992. “Reflections on Realism,” pp. 1–23 in Jock Young and Roger Matthews (eds.), Rethinking Criminology: The Realist Debate. London: Sage. McConville, Seán, 1995. “The Victorian Prison: England, 1865–1965,” in Norval Morris and David J. Rothman (eds.), The Oxford History of the Prison: The Practice of Punishment in Western Society. New York: Oxford University Press. McCorkle, Richard C., 1993. “Punish and Rehabilitate? Public Attitudes toward Six Common Crimes.” Crime and Delinquency 39:2L240–252. McDonald, Douglas, 1980. The Price of Punishment: Public Spending for Corrections in New York. Boulder, CO: Westview Press. McGowen, Randall, 1995. “The Well-Ordered Prison: England, 1780–1865,” in Norval Morris and David J. Rothman (eds.), The Oxford History of the Prison: The Practice of Punishment in Western Society. New York: Oxford University Press. McIntyre, Jennie, 1976. “Public Attitudes toward Crime and Law Enforcement,” pp. 185–203 in Robert L. Henshel and Robert A. Silverman (eds.), Perception in Criminology. New York and London: Columbia University Press. McIntyre, Jennifer, and Alissa Riker, 1993. From Beyond Shelter to Behind Bars. San Francisco: Center on Juvenile and Criminal Justice. McKelvey, Blake, 1936. American Prisons: A Study in American Social History Prior to 1915. Chicago: University of Chicago Press. McGee, Richard A., 1974a. “A New Look at Sentencing, Part I.” Federal Probation 38:2:3–8. ———, 1974b. “A New Look at Sentencing, Part II: A Plan for Maximum Justice with Minimum Capriciousness.” Federal Probation 38:3:3–11. McWilliams, Carey, 1946. Southern California: An Island on the Land. Salt Lake City, UT: Gibbs-Smith. Megargee, Edwin I., 1976. “The Prediction of Dangerous Behavior.” Criminal Justice and Behavior 3:1:3–22.
204
Bibliography
Menzies, Robert J., Christopher D. Webster, and Diana S. Sepejak, 1985. “The Dimensions of Dangerousness: Evaluating the Accuracy of Psychometric Predictions of Violence among Forensic Patients.” Law and Human Behavior 9:1:49–70. Merton, Robert K., 1938. “Social Structure and Anomie.” American Sociological Review 3:672–682. Messinger, Sheldon L., John E. Berecochea, David Rauma, and Richard A. Berk, 1985. “The Foundations of Parole in California.” Law and Society Review 19:1:69–106. Messinger, Sheldon L., and Philip D. Johnson, 1978. “California’s Determinate Sentencing Statute: History and Issues,” pp. 13–58 in Determinate Sentencing: Reform or Regression? Washington, DC: National Institute of Law Enforcement and Criminal Justice. Miethe, Terance D., 1982. “Public Consensus on Crime Seriousness: Normative Structure or Methodological Artifact?” Criminology 20:515–526. Mihata, Kevin, 1997. “The Persistence of ‘Emergence’.” pp. 30–38 in Raymond Eve, Sara Horsfall, and Mary E. Lee (eds.), Chaos, Complexity, and Sociology: Myths, Models, and Theories. Thousand Oaks, CA: Sage. Miller, Stuart J., Simon Dinitz, and John P. Conrad, 1982. Careers of the Violent: The Dangerous Offender and Criminal Justice. Lexington, MA: Lexington Books. Mitford, Jessica, 1973. Kind and Usual Punishment: The Prison Business. New York: Random. Monahan, John, 1978. “The Prediction of Violent Criminal Behavior: A Methodological Critique and Prospectus,” pp. 244–269 in Alfred Blumstein, Jacqueline Cohen, and Daniel Nagin (eds.), Deterrence and Incapacitation: Estimating the Effects of Criminal Sanctions on Crime Rates. Washington, DC: National Academy of Sciences. ———, 1981. Predicting Violent Behavior: An Assessment of Clinical Techiniques. Beverly Hills: Sage. Moore, Mark H., Susan R. Estrich, Daniel McGillis, and William Spelman, 1984. Dangerous Offenders: The Elusive Target of Justice. Cambridge: Harvard University Press. Morris, Norval, 1951. The Habitual Criminal. London: Longmans, Green. ———, 1974. The Future of Imprisonment. Chicago: University of Chicago Press. Morris, Norval, and Marc Miller, 1985. “Predictions of Dangerousness,” pp. 1–50 in Michael Tonry and Norval Morris (eds.), Crime and Justice: An Annual Review of Research. Chicago and London: University of Chicago Press. ———, 1987. Predictions of Dangerousness in the Criminal Law. Washington DC: National Institute of Justice. Morris, Norval, and David J. Rothman, 1995. “Introduction,” pp. vii–xiv in Norval Morris and David J. Rothman (eds.), The Oxford History of the Prison: The Practice of Punishment in Western Society. New York: Oxford University Press.
Bibliography
205
Myers, Martha A., 1993. “Inequality and the Punishment of Minor Offenders in the Early 20th Century.” Law and Society Review 27:2:313–343. ———, 1987. “Economic Inequality and Discrimination in Sentencing.” Social Forces 65:3:746–766. ———, 1989. “Symbolic Policy and the Sentencing of Drug Offenders.” Law and Society Review 23:2:295–315. ———, 1990. “Economic Threat and Racial Disparities in Incarceration: The Case of Postbellum Georgia.” Criminology 28:4:627–656. Nagin, Daniel, 1977. The Impact of Flat-Time Sentencing Legislation on Prison Population and Sentence Length: A California Case Study (Working Paper). Duke University: Institute of Policy Sciences and Public Affairs. ———, 1978. “General Deterrence: A Review of the Empirical Evidence,” pp. 95–139 in Alfred Blumstein, Jacqueline Cohen, and Daniel Nagin (eds.), Deterrence and Incapacitation: Estimating the Effects of Criminal Sanctions on Crime Rates. Washington DC: National Academy of Sciences. ———, 1998. “Deterrence and Incapacitation,” in Michael Tonry (ed.), The Handbook of Crime and Punishment. New York: Oxford University Press. National Council on Crime and Delinquency, 1978 (?). Instead of Prison: A Report on the Community Treatment Project for Repeat Offenders. Oakland County, MI: NCCD. National Council on Crime and Delinquency (Advisory Council of Judges), 1963. Model Sentencing Act. New York: NCCD. National Council on Crime and Delinquency (Advisory Council of Judges), 1972. Model Sentencing Act. 2d ed. New York: NCCD. National Reformatory Union, 1863. Report of the Proceedings of the Union and a Financial Statement, Together with an Abstract of the Report of the Commissioners Appointed to Inquire into the Operation of the Acts Relating to Transportation and Penal Servitude. London: National Reformatory Union [pamphlet]. National Research Council, 1993. Understanding and Preventing Violence. Albert J. Reiss and Jeffrey A. Roth (eds.). Washington, DC: National Academy Press. Office of Financial Management, State of Washington, 1986. Prison and Inmate Population Forecast, State of Washington FY 1986 to FY 1999. Olympia: Office of Financial Management, Policy Analysis and Forecasting Division. Ohlin, Lloyd E., and Frank J. Remington, 1993. Discretion in Criminal Justice: The Tension Between Individualization and Uniformity. Albany: State University of New York Press. O’Malley, Pat, 1992. “Risk, Power, and Crime Prevention.” Economy and Society 21:252–275. Owen, Barbara A., 1988. The Reproduction of Social Control: A Study of Prison Workers at San Quentin. New York: Praeger.
206
Bibliography
Packer, Herbert L., 1968, The Limits of the Criminal Sanction. Stanford: Stanford University Press. Palmer, Ted, 1992. The Re-Emergence of Correctional Intervention. Newbury Park, CA: Sage. ———, 1994. A Profile of Correctional Effectiveness and New Directions for Research. Albany: State University of New York Press. Patrick, Steven, 1991. “Organizational Control: The Empirically Driven Computer Simulation of a Sythetic Theory.” Ph.D. diss., Department of Sociology, University of California, Riverside. ———, 1995. “The Dynamic Simulation of Control and Compliance Processes in Material Organizations.” Sociological Perspectives 38:4:497–518. Pearce, Frank, and Steve Tombs, 1998. Toxic Capitalism: Corporate Crime and the Chemical Industry. Aldershot, England: Dartmouth Ashgate. Petee, Thomas, 1994. “Recommended for Release on Recognizance: Factors Affecting Pretrial Release Recommendations.” Journal of Social Psychology 134:3:375–382. Peters, Edward M., 1995. “Prison Before the Prison: The Ancient and Medieval Worlds,” in Norval Morris and David J. Rothman (eds.), The Oxford History of the Prison: The Practice of Punishment in Western Society. New York: Oxford University Press. Petersilia, Joan, 1980. “Criminal Career Research: A Review of Recent Evidence,” in Norval Morris and Michael Tonry (eds.), Crime and Justice: A Review of Research. Vol. 2. Chicago: University of Chicago Press. ———, 1986. Prison Versus Probation in California. Santa Monica: Rand. Pens, Dan, 1998. “The California Prison Guards’ Union: A Potent Political Interest Group,” pp. 134–139 in Daniel Burton-Rose, Dan Pens, and Paul Wright (eds.), The Celling of America: An Inside Look at the U.S. Prison Industry. Monroe, ME: Common Courage Press. Philips, David, 1983. “‘A Just Measure of Crime, Authority, Hunters, and Blue Locusts’: The ‘Revisionist’ Social History of Crime and the Law in Britain, 1750–1850,” pp. 50–74 in Stanley Cohen and Andrew Scull (eds.), Social Control and the State: Historical and Comparative Essays. Martin Robertson: Oxford. Pillsbury, Samuel H., 1995. “Why Are We Ignored? The Peculiar Place of Experts in the Current Debate about Crime and Justice.” Criminal Law Bulletin 31:305–336. Pisciotta, Alexander W., 1983. “Scientific Reform: The ‘New Penology’ at Elmira, 1876–1900.” Crime and Delinquency 29:613–630. Pollard, Vic, 1994. “Mike Reynolds: Profile.” CJ Weekly, March 14, 1994. Pratt, John, 1997. Governing the Dangerous: Dangerousness, Law and Social Change. Sydney: Federation Press. Pressman, Steven, and Jennifer Kaae, 1996. “Three Strikes.” California Lawyer 16:10:33–39.
Bibliography
207
Prince, Walter Franklin, 1899. “Three Studies in the History of Crime and Punishment in the Colonial Period.” Ph.D. diss., Department of History: Yale University. Proctor, Jon L., 1994. “Evaluating a Modified Version of the Federal Prison Systems’s Inmate Classification Model: An Assessment of Objectivity and Predictive Validity.” Criminal Justice and Behavior 21:2:256–272. Quetelet, Lambert Adolphe Jacques, 1842. A Treatise on Man and the Development of His Faculties. New York: Burt Franklin. Rafter, Nicole Hahn, 1997. Creating Born Criminals. Urbana: University of Illinois Press. Regional Oral History Office, Bancroft Library, University of California, Berkeley, 1985. Law Enforcement and Criminal Justice: An Oral History Conducted 1966–1974. Berkeley: University of California. Reichman, Nancy, 1986. “Managing Crime Risks: Toward and Insurance Based Model of Social Control.” Research in Law, Deviance, and Social Control 8:151–72. Rennie, Ysabel, 1978. The Search for Criminal Man: A Conceptual History of the Dangerous Offender. Lexington, MA: Heath. Rigakos, George S., 1999. “Risk Society and Actuarial Criminology: Prospects for a Critical Discourse.” Canadian Journal of Criminology 41:2:137–152. Roemer, John, 1996. “Taking Three Strikes Outside the Walls.” California Lawyer 16:10:40–41/83–87. Rossi, Peter H., Jon E. Simpson, and JoAnn L. Miller, 1985. “Beyond Crime Seriousness: Fitting the Punishment to the Crime.” Journal of Quantitative Criminology 1:1:59–90. Roth, Martin, 1993. The Writer’s Complete Crime Reference Book. Cincinnatti: Writer’s Digest. Rothman, David J., 1971. The Discovery of the Asylum: Social Order and Disorder in the New Republic. Boston: Little, Brown. ———, 1980. Conscience and Convenience: The Asylum and Its Alternatives in Progressive America. Boston: Little, Brown. ———, 1983. “Sentencing Reforms in Historical Perspective.” Crime and Delinquency 29:4:631–647. ———, 1995. “Perfecting the Prison: United States, 1789–1865,” in Norval Morris and David J. Rothman (eds.), The Oxford History of the Prison: The Practice of Punishment in Western Society. New York: Oxford University Press. Rotman, Edgardo, 1990. Beyond Punishment: A New View on the Rehabilitation of Criminal Offenders. New York: Greenwood Press. Rusche, Georg, and Otto Kirscheimer, 1939. Punishment and Social Structure. New York: Columbia University Press. Russell, Bertrand, 1950. Unpopular Essays. London: Allen and Unwin.
208
Bibliography
Sampson, Robert J., and John Laub, 1993. “Structural Variations in Juvenile Court Processing: Inequality, the Underclass, and Social Control.” Law and Society Review 27:2:285–311. Sarbin, Theodore R., 1967. “The Dangerous Individual: An Outcome of Social Identity Transformations.” British Journal of Criminology 7:3:285–295. Schelling, Thomas C., 1978. Micromotives and Macrobehavior. New York: Norton. Schiraldi, Vincent, 1994. The Undue Influence of California’s Prison Guards’ Union: California’s Correctional-Industrial Complex. San Francisco: Center on Juvenile and Criminal Justice. Schiraldi, Vincent, with Michael Godfrey, 1994. Racial Disparities in the Charging of Los Angeles County’s “Third Strike” Cases. San Francisco: Center on Juvenile and Criminal Justice. Schrag, Peter, 1998. Paradise Lost: California’s Experience, America’s Future. Berkeley: University of California Press. Schrager, Laura Shill, and James F. Short Jr., 1980. “How Serious a Crime? Perceptions of Organizational and Common Crimes,” pp. 14–31 in Gilbert Geis and Ezra Stotland (eds.), White Collar Crime: Theory and Research. Sage Criminal Justice System Annuals. Vol. 13. Beverly Hills, CA: Sage. Schulhofer, Stephen J., 1993. “Rethinking Mandatory Minimums.” Wake Forest Law Review 28:199–222. Scott, P. D., 1977. “Assessing Dangerousness in Criminals.” British Journal of Psychiatry 131:124–142. Scott, S. P., 1932. The Civil Law, Including the Twelve Tables, the Institutes of Gaius, the Rules of Ulpian, the Opinions of Paulus, the Enactments of Justinian, and the Constitutions of Leo. Vols. 1, 2, and 7. Translated from the original Latin, edited, and compared with all accessible systems of jurisprudence ancient and modern. Cincinnati: Central Trust Company. Sellin, Thorsten, 1938. Culture Conflict and Crime. New York: Social Science Research Council. ———, 1976. Slavery and the Penal System. New York: Oxford. Shah, Saleem A., 1981. “Legal and Mental Health System Interactions: Major Developments and Research Needs.” International Journal of Law and Psychiatry 4:3-sup4:219–270. Shane–DuBow, Sandra, Alice P. Brown, and Erik Olsen. Sentencing Reform in the United States: History, Content, and Effect. Washington, DC: National Institute of Justice. Shannon, Lyle W., 1991. Changing Patterns of Delinquency and Crime: A Longitudinal Study in Racine. Boulder: Westview Press. Shaw, A.G.L., 1966. Convicts and the Colonies: A Study of Penal Transportation from Great Britain and Ireland to Australia and Other Parts of the British Empire. London: Faber.
Bibliography
209
Sherman, Michael, and Gordon Hawkins, 1981. Imprisonment in America: Choosing the Future. Chicago and London: University of Chicago Press. Shichor, David, 1987. “Penal Policies: Some Recent Trends.” Legal Studies Forum 11:55–78. ———, 1997. “Three Strikes as Public Policy: The Convergence of the New Penology and the McDonaldization of Punishment.” Crime and Delinquency 43:4:470–492. Shinnar, Shlomo, and Reuel Shinnar, 1975. “The Effects of the Criminal Justice System on the Control of Crime: A Quantitative Approach.” Law and Society Review 9:581–611. Simon, Jonathan, 1987. “The Emergence of a Risk Society: Insurance, Law, and the State.” Socialist Review 95:61–89. ———, 1993. Poor Discipline: Parole and the Social Control of the Underclass, 1890–1990. Chicago and London: University of Chicago Press. Sklansky, David A., 1995. “Cocaine, Race, and Equal Protection.” Stanford Law Review 47:1283–1322. Sleffel, Linda, 1977. The Law and the Dangerous Criminal: Statutory Attempts at Definition and Control. Lexington, MA: Lexington Books. Smith, Douglas A., 1986. “The Neighborhood Context of Police Behavior,” pp. 313–342 in Albert J. Reiss Jr., and Michael Tonry (eds.), Communities and Crime. Crime and Justice: An Annual Review of Research. Chicago: University of Chicago Press. Spelman, William, 1994. Criminal Incapacitation. New York: Plenum. Spierenburg, Pieter (ed.), 1984. The Emergence of Carceral Institutions: Prisons, Galleys, and Lunatic Asylums 1550–1900. Rotterdam, Netherlands: Erasmus Universiteit. ———, 1991. The Prison Experience: Disciplinary Institutions and Their Inmates in Early Modern Europe. New Brunswick: Rutgers University Press. ———, 1995. “The Body and the State: Early Modern Europe,” in Norval Morris and David J. Rothman (eds.), The Oxford History of the Prison: The Practice of Punishment in Western Society. New York: Oxford University Press. Starr, M. B., 1873. The Coming Struggle: What the People on the Pacific Coast Think of the Coolie Invasion. San Francisco: Bacon & Company. State Board of Prison Directors, 1910. Report of the State Board of Prison Directors, State of California, on a Proposed Reformatory for Adult Offenders. Sacramento: W. W. Shannon, Superintendent State Printing. Steadman, Henry J., and Joseph J. Cocozza, 1974. Careers of the Criminally Insane. Lexington, MA: Heath. Steury, Ellen Hochstedler, and Nancy Frank, 1990. “Gender Bias and Pretrial Release: More Pieces of the Puzzle.” Journal of Criminal Justice 18:417–432.
210
Bibliography
Strachan-Davidson, James Leigh, 1912. Problems of the Roman Criminal Law. Oxford: Clarendon Press. Stolzenberg, Lisa, and Stewart J. D’Alessio, 1997. “ ‘Three Strikes and You’re Out’: The Impact of California’s New Mandatory Sentencing Law on Serious Crime Rates.” Crime and Delinquency 43:4:457–469. Sutherland, Edwin H., 1947. Principles of Criminology. 4th ed. Chicago: Lippincott. Tallack, William, 1889. Penological and Preventive Principles. London: Wertheimer, Lea. Tappan, Paul W., 1949. “Habitual Offender Laws in the United States.” Federal Probation 13:1:28–31. Tarde, Gabriel, 1903. The Laws of Imitation. Translated by Elsie Clews Parsons. London: William Heinemann. ———, 1912. Penal Philosophy. Translated by Rapelje Howell. London: William Heinemann. Tarling, Roger, and John A. Perry, 1985. “Statistical Methods in Criminological Prediction,” pp. 210–231 in David P. Farrington and Roger Tarling (eds.), Prediction in Criminology. Albany: State University of New York Press. Taylor, Lawrence, 1984. Born to Crime: The Genetic Causes of Criminal Behavior. Westport, CT: Greenwood Press. Teller, Paul, 1992. “A Contemporary Look at Emergence,” pp. 139–153 in Ansgar Beckermann, Hans Flohr, and Jaegwon Kim (eds.), Emergence or Reduction? Essays on the Prospects of Nonreductive Physicalism. Berlin and New York: Walter de Gruyter. Thornberry, Terence P. and Joseph E. Jacoby, 1979. The Criminally Insane: A Community Follow–up of Mentally Ill Offenders. Chicago: University of Chicago Press. Tonry, Michael, 1994. “Sentencing Guidelines, Disadvantaged Offenders, and Racial Disparities.” Philosophy and Public Policy 14:3:7–13. ———, 1995. Malign Neglect: Race, Crime, and Punishment in America. New York: Oxford University Press. Tonry, Michael, and Norval Morris, 1978. “Sentencing Reform in America,” pp. 249–266 in Gordon Hawkins and Franklin E. Zimring, 1984 (eds.), The Pursuit of Criminal Justice: Essays from the Chicago Center. Chicago and London: University of Chicago Press. Trebach, Arnold S., 1987. The Great Drug War, and Radical Proposals that Could Make America Safe Again. New York: Macmillan. Turk, Austin T., 1969. Criminality and Legal Order. Chicago: Rand-McNally. ———, 1976. “Law as a Weapon in Social Conflict.” Social Problems 23:3:276–291. ———, 1982. Political Criminality: The Defiance and Defense of Authority. Newbury Park, CA: Sage.
Bibliography
211
Turner, Jonathan H., 1998. “Must Sociological Theory and Sociological Practice Be so Far Apart? A Polemical Answer.” Sociological Perspectives 41:2:243–258. Turner, Susan, 1999. “Profiling Inmates in the Los Angeles County Jail: Risks, Recidivism, and Release Options.” Paper presented at National Institute of Justice Annual Conference on Criminal Justice Research and Evaluation, July 18–21, 1999, Washington, DC. Ulmer, Jeffery T., 1997. Social Worlds of Sentencing: Court Communities Under Sentencing Guidelines. Albany: State University of New York Press. Unitarian Universalist Service Committee, 1982. Alternatives to Imprisonment: A Thoughtful Approach to Crime and Punishment. Boston: UUSC. U.S. Department of Justice, Bureau of Justice Statistics, 1993. Survey of Inmates of State Correctional Facilities 1991 [Computer file]. Conducted by U.S. Dept. of Commerce, Bureau of the Census. ICPSR ed. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [producer and distributor], 1993. U.S. Department of Justice, 1994. An Analysis of Non-Violent Drug Offenders with Minimal Criminal Histories. Washington, DC: U.S. Department of Justice. van den Haag, Ernest, 1975. Punishing Criminals: Concerning a Very Old and Painful Question. New York: Basic. van Dijk, Jan J. M., 1989. “Penal Sanctions and the Process of Civilization.” Annales Internationales de Criminologie 27:191–204. Van Dine, Stephan, John P. Conrad, and Simon Dinitz, 1979. Restraining the Wicked: The Incapacitation of the Dangerous Criminal. Lexington, MA: Lexington Books. Visher, Christy A., 1986. “The Rand Inmate Survey: A Reanalysis,” in Alfred Blumstein, Jacqueline Cohen, Jeffrey A. Roth, and Christy A. Visher (eds.), Criminal Careers and “Career Criminals.” Vol. 2. Washington, DC: National Academy Press. Vold, George B., 1958. Theoretical Criminology. New York: Oxford University Press. von Hirsch, Andrew, 1976. Doing Justice: The Choice of Punishments. New York: Hill and Wang. ———, 1984. “The Ethics Of Selective Incapacitation: Observations on the Contemporary Debate.” Crime and Delinquency 30:175–194. ———, 1985. Past or Future Crimes: Deservedness and Dangerousness in the Sentencing of Criminals. New Brunswick: Rutgers University Press. ———, 1993. Censure and Sanctions. Oxford: Clarendon Press. von Hirsch, Andrew, and Andrew Ashworth (eds.), 1992. Principled Sentencing. Boston: Northeastern University Press. von Hirsch, Andrew, and Don M. Gottfredson, 1984. “Selective Incapacitation: Some Queries on Research Design and Equity.” Review of Law and Social Change 12:11–51.
212
Bibliography
von Hirsch, Andrew, and Kathleen Hanrahan, 1981. “Determinate Penalty Systems in America: An Overview. Crime and Delinquency 27:3:289–316. von Hirsch, Andrew, and Julia M. Mueller, 1984. “California’s Determinate Sentencing Law: An Analysis of its Structure.” New England Journal of Civil and Criminal Confinement 10:2:253–300. Walker, Nigel, 1978. “Dangerous People.” International Journal of Law and Psychiatry 1:1:37–50. ———, 1991. Why Punish? Theories of Punishment Reassessed. Oxford: Oxford University Press. ———, 1994. “Dangerousness and Mental Disorder,” pp. 179–190 in A. Phillips Griffiths (ed.), Philosophy, Psychology, and Psychiatry. Cambridge: Cambridge University Press. Warr, Mark, 1984. “Fear of Victimization: Why Are Women and the Elderly More Afraid?” Social Science Quarterly 65:3:681–702. Warr, Mark, and Mark Stafford, 1983. “Fear of Victimization: A Look at the Proximate Causes.” Social Forces 61:4:1033–1043. Webster, Christopher, Bernard Dickens, and Susan Addario, 1985. Constructing Dangerousness: Scientific, Legal, and Policy Implications. Toronto: Centre of Criminology, University of Toronto. Webster, Christopher, Grant T. Harris, Marnie E. Rice, Catherine Cormier, and Vernon L. Quinsey, 1994. The Violence Prediction Scheme: Assessing Dangerousness in High Risk Men. Toronto: Centre of Criminology, University of Toronto. Weidenhofer, Margaret, 1973. The Convict Years: Transportation and the Penal System 1788–1868. Melbourne: Lansdowne Press. Wenk, Ernst. A., James O. Robison, and Gerald W. Smith, 1972. “Can Violence Be Predicted?” Crime and Delinquency 18:4:393–402. Whyte, Edward H., 1916. “The Parole System in California” 7 Journal of the American Institute of Criminal Law and Criminology 1. Wicharaya, Tamasak, 1995. Simple Theory, Hard Reality: The Impact of Sentencing Reforms on Courts, Prisons, and Crime. Albany: State University of New York Press. Wildavsky, Aaron B., 1979. Speaking Truth to Power: The Art and Craft of Policy Analysis. Boston: Little, Brown. Wilkins, Leslie T., 1985. “The Politics of Prediction,” pp. 34–51 in David P. Farrington and Roger Tarling, 1985 (eds.), Prediction in Criminology. Albany: State University of New York Press. Wilson, James Q., 1975. Thinking about Crime. New York: Random. ———, 1983. Thinking about Crime. 2nd ed. New York: Basic.
Bibliography
213
Wilson, James Q., and Richard J. Herrnstein, 1985. Crime and Human Nature. New York: Simon & Schuster. Wilson, William Julius, 1978. The Declining Significance of Race: Blacks and Changing American Institutions. Chicago: University of Chicago Press. Wines, Enoch C. (ed.), 1871. Transactions of the National Congress on Penitentiary and Reformatory Discipline Held at Cincinatti, Ohio, October 12–18, 1870. Albany: Weed, Parsons. Wolfgang, Marvin E., Robert M. Figlio, and Thorsten Sellin, 1972. Delinquency in a Birth Cohort. Chicago: University of Chicago Press. Wolfgang, Marvin E., Robert M. Figlio, Paul E. Tracy, and Simon I. Singer, 1985. The National Survey of Crime Severity. U.S. Department of Justice, Bureau of Justice Statistics. Washington, DC: USGPO. Woznicki, Robert, 1968. History of Yuma and the Territorial Prison. Calexico, CA: San Diego State College. Wright, James D., and Rossi, Peter H., 1986. Armed and Considered Dangerous: A Survey of Felons and Their Firearms. New York: Aldine de Gruyter. Young, Jock, 1986. “The Failure of Criminology: The Need for a Radical Realism,” pp. 4–30 in Roger Matthews and Jock Young (eds.), Confronting Crime. London: Sage. ———, 1992. “Ten Points of Realism,” pp. 24–68 in Jock Young and Roger Matthews (eds.), Rethinking Criminology: The Realist Debate. London: Sage. Zalman, Marvin, 1977–1978. “The Rise and Fall of the Indeterminate Sentence.” Wayne Law Review 24:1, 24:3. Zedlewski, Edwin W., 1987. Making Confinement Decisions. Washington, DC: National Institute of Justice. Zimbardo, Philip G., 1994. Transforming California’s Prisons into Expensive Old Age Homes for Felons: Enormous Hidden Costs for California’s Taxpayer. San Francisco: Center on Juvenile and Criminal Justice. Zimring, Franklin E, 1981. “Sentencing Reform in the States: Lessons from the 1970s,” pp. 101–121 in Michael Tonry and Franklin E. Zimring, 1983 (eds.), Reform and Punishment: Essays on Criminal Sentencing. Chicago and London: University of Chicago Press. ———, 1983. “Sentencing Reform in the States: Lessons from the 1970s,” pp. 101–121 in Michael Tonry and Franklin E. Zimring (eds.), Reform and Punishment: Essays on Criminal Sentencing. Chicago and London: University of Chicago Press. Zimring, Franklin E., and Gordon Hawkins, 1973. Deterrence: The Legal Threat in Crime Control. Chicago: University of Chicago Press. ———, 1988. “The New Mathematics of Imprisonment.” Crime and Delinquency 34:4:425–436.
214
Bibliography
———, 1991. The Scale of Imprisonment. Chicago: University of Chicago Press. ———, 1994. “The Growth of Imprisonment in California.” British Journal of Criminology 34:83–96. ———, 1995. Incapacitation: Penal Confinment and the Restraint of Crime. New York: Oxford Univeristy Press. ———, 1997. Crime Is Not the Problem: Lethal Violence in America. New York: Oxford University Press. Zimring, Franklin E., Gordon Hawkins, and Sam Kamin, 2001. Punishment and Democracy: Three Strikes and You’re Out in California. New York: Oxford University Press.
Subject Index
See also Arrest; See also Ballot initiatives; See also CCPOA See also Determinate Sentencing Law; See also Jail; See also Prisons, California; See also Probation; See also Parole CANJUS, 10 Capital punishment, See Punishment, capital Career criminals, 7–8, 10, 35, 55, 62–63, 65–66, 69–70, 77, 79 80, 88, 181 n11, 182 n9 CCPOA, (California Correctional Peace Officers Association), 43 Classical criminology, 21–22, 27, 34 Classification, 36, 46, 62 passim, 65–66, 80–86, 94, 105 Corporal punishment, See Punishment, corporal Criminal careers, See Career criminals Criminal punishment, See Punishment, criminal
Actuarial criminology, 7, 36, 176 n36 Actuarialism, See Actuarial criminology Age, Cost of housing inmates, and 56 Crime and, 27, 56, 62–63, 87–88 Dangerousness and, 78, 85, 91–92, 97, 108–109, 110, 114, 116–117, 119–120, 128, 133, 137–138, 141–142, 148, 155, 157 See also Geriatric release African-Americans, see race Analytic Realism, 13–15, 156–157 Arrest, As measure of criminality 67, 69, 82, 84, 92–93 As part of criminal justice system 11, 127–128 Population (California) Dangerousness of 108–112, 127–128 Size of 105 Auburn State Penitentiary, 28 Australia, See Transportation Ballot initiatives, 5, 40–43, 47–48, 51–52, 54–57, 177 n3, n5, 178 n 13, 179 n17 Ballot Propositions, See Ballot initiatives Baxstrom v. Herold, 81–82 Boatwright Amendment, 48, 177 n14 Burgess, Ernest W., 84, 91, 108 California, History 6–8, 39–58
Dangerous offenders, Defining 6–8, 73–74, 76–80, 90–93, 157, 181 n7, 182 n9 Identification of, 8, 80, 84–86, 46, 67–68 Incapacitation of, 6, 13, 15–16, 36, 40, 54, 58, 72–73, 87–90, 98, 100, 103, 113, 128–129, 131, 149, 151–155, 157–158, 172 n11; See also Career criminals
215
216
Subject Index
Dangerousness, Conceptualizing, 7–9, 15, 58 passim, 72–86, 108, 171 n5, 178 n15 passim, 181 n2, 182 n8 Measuring, 9, 11, 13, 15, 72 passim, 87–100, 172 n9, n11, 180 n7 passim, 182 n3, n4, 184 n6, 185 n2, In California criminal justice system populations 13, 15, 89, 108, 127–128 See also Individual criminal justice system populations Death penalty, See Punishment, capital Determinate Sentencing Law (California), 40, 48–51, 58 Deterrence, 17–19, 21–25, 28, 32–34, 43, 172 n11, 176 n11, 186 n9 Displacement (of offenders), 116, 129 Dixon et al. v. Attorney General of the Commonwealth of Pennsylvania, 81–82 Drug offenders, 94 113, 116, 119, 123, 128–129, 132, 155, 157, 183 n8 Drug policy, 33–34, 47, 51, 57, 67, 71, 84–85, 87, 113, 128–129, 132, 155, 157, 172 n2, 182 n1 Dynamic systems modeling, See Modeling, dynamic systems Emergence, 107 England, 19–22, 25, 80, 174 n19 Evaluation, 9, 15, 71–73, 86, 98, 128, 154–155, 181 n1 False positives, See Prediction, errors in Folsom Prison, 40, 44, 46, 53 Forrester, Jay, 11, 15, 153–154, 183 n9 Gender, See Sex Geriatric release, 132, 144–149, 151, 154–155, 165, 185 n5 n6, 186 n8 Great Britain, See England Habitual offenders, See Career criminals Hispanic, See Race Howard, John, 26
Imprisonment, See Prisons Incapacitation, As justification for punishment, 4, 7, 17, 31–37, 39, 44, 80 Collective, 61 Selective, Actuarial criminology and, 36, 176 n36 Definition 4, 8, 35, 58, 61, 176 n34 passim, Evaluation of, 8–9, 15, 35, 58, 68–73, 75, 86, 90–91, 97–98, 123, 127–129, 148, 151, 154–155, 157, 181 n10 In California 40, 54–56, 178 n15m 178–179 n17, Problems in implementing, 63–68, 84–85, 180 n7, Prominence in current policy discourse 6–8, 35–36, 62, 90 Social defense and, 8, 57, 75, 129, 157 Incarceration, See Prisons Independence of observations (in modeling), 97 Indeterminate sentencing, See Sentencing, indeterminate Jail, As part of criminal justice system, 7, 10–11, 13, 15, 70 passim, 87, 95–96, 98, 129 Population (California) Dangerousness of, 13, 109, 112–115, 127–128 Size of, 58, 88, 106, 129 JUSSIM, 10 Lambda (), 8, 68–71 Legal Dangerousness Scale (Steadman and Cocozza), 92 Left Realism, See Realist criminology Mandatory minimum sentences, See Sentences, Mandatory minimum McGee, Richard, 47–48, 171n
Subject Index Modeling, Dynamic systems 9–13, 90, 95, 97–98, 103, 105, 131, 160–161, 171 n6, 172 n8, 184 n1, Statistical, 97, 183 n13 Simulation, 11–13, 85, 94–99, 103–105, 131, 149 passim, 153, 155, 159–162, 167–169,183 n13, 186 n8 passim, Paradigm, Definition, 17–18 In criminal justice 4, 6, 19, 21, 23–25, 27–29, 31–34, 36–37, 39–40, 46–48, 80, 172 n2, n4, 174 n17, 175 n23, 176 n34 passim, 177 n38, 178 n17 Shifts, 17–18, 25–26, 30–31, 36, 39–43, 45, 49, 52, 58, 98 passim, 171 n1, 171 n4, 172 n3, 173 n13, 175 n25, 176 n36, 177 n2 See also Deterrence; See also Incapacitation; See also Rehabilitation; See also Retribution Parole, As part of criminal justice system, 13, 15, 29, 42, 44–51, 56–57, 95–96, 129, 178 n10 Population (California), Dangerousness of, 90, 99, 123–129 Size of, 88, 107 Use of prediction in, 67, 80, 84–86 Penitentiary, History, 25–28, 174 n19 Purpose, 24–27, 28, 44, 173 n11 People v. Jackson, 179 n18 People v. Ramirez, 179 n18 Policy evaluation, See Evaluation Positivist criminology, 27, 34, 174–175 n20 Prediction, Actuarial, See statistical Anamnestic, 63, 77 Clinical, 63, 80–82, 182 n10, Criminal justice decisionmaking and, 9, 35, 62, 66–67, 76, 80–86, 89, 180 n7, 182 n10
217
Errors in, 63–67, 80–81, 89, 180 n6, 180 n7 Objective instruments for 84–86 Simulation and, relationship between, 98–99 Statistical, 63, 65, 78, 80, 180 n4 See also Preventive detention, See also Predictive evaluation Predictive evaluation, 13, 99, 131–152 Preventive detention, 62, 66–67, 80 Prisons, As part of criminal justice system, 4, 10–11, 15, 17, 95–98, 172 n1, California, Costs of, 52, 54–55, 88, 154, 156 Facilities, 43–49, 52–53, 88, 178 n8 See also individual names of facilities Population, 5, 45, 49–50, 52, 56–57, 72, 88, 107, 132 Increase in, 6–7, 33, 35, 37, 49, 56–57, 88, 122, 129 Consequences of future policy scenarios to, 132–151, 154 Dangerousness of, 9, 13, 15, 55–56 passim, 72, 76, 90, 94 passim, 99, 116, 119–123, 125, 127–129, 154–156, 179 n17 See also Three Strikes; See also CCPOA Capacity limits, See Crowding Crowding, 3, 6, 41–42, 45, 52–53, 58, 87–88, 113, 176 n31, 195 n10 passim History and development of, 17, 19–21, 23, 26–28, 40, 44–45, 175 n23 Overcrowding See Crowding Purpose of, 3, 18–19, 23, 25–32, 34, 36–37, 44–46, 49, 52, 93–94, 172 n3, 173 n11, 177 n39, 178 n9, 179 n1, 183 n11 passim Population, United States, 71, 157 Industry, 43 See also Sentencing; See also Incapacitation; See also Classification
218
Subject Index
Probation, As part of criminal justice system, 13, 15, 51, 70, 129 Population (California), Dangerousness of, 99, 113, 116–118, 127–129, 179 n21 passim Size of, 57, 88, 106, 129 Property offenders, 92–93, 113, 116, 123, 129, 132, 141 Proposition 184, See Three Strikes Public opinion, 36, 41, 156, 177 n2 Public policy, 5, 16, 153–158, 172n, 176n, 183n See also Evaluation Punishment, Capital, 22, 27, 51, 83, 173 n6, See also Texas, Death penalty in Corporal, 19–24, 26–27, 173 n13, 174 n14 Disparity in, 5, 31–32, 40, 42, 49, 123, 183n11 See also Race History of, 17–37 Purpose of, 3–6, 14–15, 17–37, 41, 43, 49, 52, 66–67, 73, 173 n5, n7, n12, 177 n39, 181 n5 See also Deterrence; See also Incapacitation; See also Rehabilitation; See also Retribution; See also Penitentiary; See also Transportation Quakers, 27 Race, As a predictive characteristic, 62–63, 68, 92–93, 172n9, 182 n4, Effect on criminal justice system processing, 5, 31, 97, 105 passim, 136, 140, 144, 148, 172n9, 183 n11, 186 n2, See also Punishment, disparity Radical Realism, See Realist criminology Reagan, Ronald, 48, 178 n12 Realist criminology, 13–15, 156–158 Recidivism, 20, 77, 79, 94, 175 n24 See also Career criminals Referendum, See Ballot initiative
Rehabilitation, 4, 7, 17–19, 23–24, 27–32, 34, 36–37, 39–40, 42–48, 52, 80, 155, 171 n4, 172 n3, 173 n13, 174 n17, 175 n23, 176 n28 n35, 177 n39 Retribution, 17–19, 24–25, 31–34, 36–37, 40, 43, 49, 51, 56, 173 n5 n6 n34, 178 n9, 186 n9 Risk, 7, 63–64, 67, 73–76, 78, 85–86, 94, 113, 153, 181 n2 n4, 183 n7 San Quentin, 40, 44, 46, 53 Selective incapacitation, See Incapacitation, selective Sentencing, Determinate, 31–33, 40, 43, 48–50, 52, 58, 175 n26 n27, 176 n28, 178 n10, 183 n11 Indeterminate, 29–30, 39–40, 42, 46–50, 177 n1 Mandatory, 33, 51–52, 56, 87, 89, 95, 113, 132, 154, 164, 178 n10, 183 n8 Habitual offender, 5, 7, 10, 36, 39, 55–56, 62, 87–88, 132 See also Three Strikes; See also Punishment Separate System, 28, 175 n21 Sex, As indicator of dangerousness, 11, 108, 192 n4 See also Women, incarceration of Silent System, 28, 175 n21 Simulation modeling See Modeling, simulation Social defense, 15, 18, 20, 31, 33, 36, 45–46, 75–76, 157, 172 n4, 176 n34, 185 n5 See also Utilitarianism; See also Incapacitation, selective Statistics, limitations of, 68, 97, 99, 131, 156 See also Modeling, statistical System analysis, See Modeling, dynamic systems Systems modeling, See Modeling, dynamic systems
Subject Index Texas, death penalty in, 83–84 Three Strikes (and You’re Out), Effects on prison population, 88–89, 121, 123, 125, 154, 183 n11 Evaluating, 6, 13, 88–89, 99, 123 passim, 131–154 In California, 5, 15, 39–40, 56–58, 87, 182 n1 History, 40–41, 43, 55–56, 177 n4, 178–79 n17 In different states, 56, 179 n19, 182 n1 passim Selective incapacitation and, 6, 10, 36, 56–58, 62, 87, 154, 178–79 n17, 186 n9 Systemic effects of, 58, 87, 98–99, 113, 131, 154
219
Transportation, 22, 24, 25–26, 173 n12 Truth in Sentencing, 6, 57, 87, 89, 185 n5 Utilitarianism, 18–19, 31–32 Victims’ rights, 41, 43, 48, 51–52, 55 Violent offenders, 68, 86, 88, 90, 105, 116, 121, 129, 132, 135, 140–141, 149, 151, 161 Wabau , 178 n8 Walnut Street Jail, 37 Women, incarceration of, 53, 92, 121–123, 185 n10
This page intentionally left blank.
Author Index
Abrahamse, Allan, 35, 61–62, 64–69, 72, 84, 89–90, 180 n6, 181 n10 n11 Alexander, Jack, 80, 85 Allen, Francis A., 7, 18, 26, 19, 32, 175 n27 Allen, Hilary, 78 Allswang, John M., 41 Ancel, Marc, 76 Applegate, Brandon K., 156 Ashman, Allan, 51 Ashworth, Andrew, 6 Assembly Committee on Criminal Justice (California), 51–52 Auerhahn, Kathleen, 35, 44, 64–66, 68, 72, 75, 85, 89, 99, 145 Austin, James, 5, 14, 49, 56–57, 62–63, 76, 85–88, 132, 157, 182 n1 Avi-Itzhak, Benjamin, 181 n10 Axelrod, Robert, 98 Axtell, Robert, 97 Bales, William D., 4, 15, 87–88, 90, 99 Baum, Dan, 34 Baum, Noah, 5 Bean,Walton, 44 Beccaria, Cesare, 22 Beck, Allen J., 5, 88 Beck, James L., 84–85 Beck, Ulrich, 7 Bedrick, Brooke, 5 Beirne, Piers, 27, 34, 174 n20 Belkin, Jacob, 10 Benekos, Peter J., 62 Berk, Richard A., 51 Bertram, Eva, 34, 172 n2
Black, Donald, 93 Blackmore, John, 64, 68 Bloom, Barbara, 92, 121 Boies, John L., 97 Bookspan, Shelley, 43–45, 48, 178 n9 Boucher, R. J., 78 Blumstein, Alfred, 7–8, 10, 33, 65–66, 68–70, 78, 92–93, 95, 176 n29, 182 n4 Boswell,Terry E., 44 Brackman, Harold, 51 Braithwaite, John, 14, 157, 176 n36 Brereton, David, 39, 50 Brewer, David, 49–50 Bridges, George S., 97, 172 n8 Briggs, John, 20–21, 173 n10 Bronson, Richard, 12, 98 Brown, M. Craig, 75 Buchanan, Robert A., 85 Bunn, Ivan, 78 Bureau of Justice Statistics, 50, 88, 92 Byrne, David, 98–99 California Department of Corrections, 43, 47, 53, 88, 95 California Department of Justice, 5, 88, 129, 159 California District Attorneys Association, 49 Canela-Cacho, José A., 69–70, 99 Cantor, Nathaniel F., 23 Carey,William T., 49–50 Cary, Seth C., 26 Carlson, James M., 77 Casper, Jonathan D., 39, 50,
221
222
Author Index
Cassidy, R. Gordon, 10–11, 95 Cassou, April Kestell, 32 Caulfield, S. R., 36–37 Caulkins, Jonathan P., 33–34 Chaiken, Marcia R., 36, 62, 66, 69, 76, 90, 179 n3 Chaiken Jan M., 36, 62, 66, 69, 76, 90, 179 n3 Chapin, Bradley, 22–23 Chesney-Lind, Meda, 66, 92, 121 Childress, Michael, 171 n6 Chiricos, Ted, 97 Christie, Nils, 43 Clark, John, 56–57, 62 Clear,Todd R., 4–6, 31, 34, 171 n3 Clinard, Marshall B., 14, 75 Cocozza, Joseph J., 35, 65, 78–79, 81–83, 89–90, 92 Cohen, Jacqueline, 10, 35, 61, 64, 66–70 Cohen, Lawrence E., 32 Cohen, Stanley, 158 Cohn, Ellen G., 176 n32 Collins, Randall, 98, Congressional Information Service, 4, 30 Connolly, Kathleen, 5 Conrad, John P., 76, 80 Conte, Rosaria, 99 Copas, John, 9, 65, 68, Corrections Digest, 50 Costello, Anthony, 95, 99 Coulson, Grant, 85 Crawford, Charles, 97 Crutchfield, J. P., 98 Crutchfield, Robert D., 97 Cullen, Francis T., 156, 176 n35 Currie, Elliott, 14–15 D’Alessio, Stewart J., 5 Dahrendorf, Ralf, 14 Davies, J., 79 Davis, Christopher, 5, 136, 183 n11 Davis, Mike, 43, 52 Dees, Linda G., 4, 15, 87–88, 90, 99 Dershowitz, Alan M., 4, 29, 66 DeWolf, L. Harold, 23, 174 n19
Dinitz, Simon, 80 Dombey–Moore, Bonnie, 171 n6 Dorny, C. Nelson, 96–97 Douglas, Mary, 7 Durkheim, Emile, 13–14, 173 n5, 181 n1 Ekland-Olson, Sheldon, 84 Elias, Norbert, 23, 174 n15 n16 Ellingwood, Herbert, 48 Engelhardt–Greer, Simone , 37 Epstein, Joshua M., 97 Estes, Richard, 136, 183 n11 Erikson, Kai T., 78 Farrington, David P., 78, 92–93, 145, 176 n32 Feeley, Malcolm M., 3, 7, 36 Felson, Marcus, 32 Fernandez, Kenneth E., 85–86 Ferraro, Kenneth F., 94, 183 n7, Figlio, Robert M., ix, 35, 62–63, 69, 92, 161 Fisher, Bonnie S., 156 Flanagan, Timothy J., 29, 36–37, 156, 177 n39 Floud, Jean, 74–76, 80 Foote, Caleb, 40 Forrester, Jay W., 11, 15, 95–96, 153–154, 183 n9 Foster, Michele, 5 Foucault, Michel, 4, 21, 23–24, 30, 74, 173 n13, 174 n14 Frank, Nancy, 161 Frankel, Marvin E., 30 Friedman, Lawrence M., 21–23, 173 n9 n13, 175 n21 Friedrich, Barbara, 42 Garland, David, 23, 41, 172 n3, 173 n12 Garnett, Rick, 95, 99 Garofalo, R. F., 78 Geis, Gilbert, 75, 78 Gerber, Jurg, 37 Gibbs, Jack P., 176 n29 Gilbert, Karen E., 176 n35
Author Index Gilbert, Nigel, 97, 99 Gillard, Darrell K., 5, 88 Gilmore, Ruth Wilson, 43 Glaser, Daniel, 47, 171 n4, 178 n12 Godfrey, Michael, 5, 136 Goodstein, Lynne, 4 Gordon, Diana R., 34 Gordon, Robert A., 182 n2 Gorton, Joe, 97 Gottfredson, Don M., 36, 62–64, 67–68,85 Gottfredson, Michael R., 34–35, 77 Gottfredson, Stephen D., 62–64, 67–68, 80, 85 Graddy, Elizabeth, 93 Greenberg, David F., 35, 61 Greenfield, Lawrence A., 92 Greenwood, Peter W., 5, 13, 35, 61–62, 64–69, 72, 80, 84, 89–90, 180 n6, 181n10 n11 Griset, Pamala L., 37, 72, Guerin-Gonzales, Camille, 44 Gusfield, Joseph R., 75 Hakeem, Michael, 84 Hale, Chris, 183n Hamparian, Donna Martin, 80 Hanneman, Robert A., ix, 11, 95–96, 98, 171 n6, 183 n9 Hanrahan, Kathleen, 31–33 Harring, Sidney L., 75 Hart, H.L.A., 171 n2 Hawkins, Gordon, 5–8, 13, 34–36, 67–68, 70–71, 94, 157, 171 n2 n3, 176 n29, 181 n12 Henry, D. Alan, 56, 62 Hepburn, John R., 4 Herrnstein, Richard J., 34 Hewitt, John D., 4, 31, 34 Hibbert, Christopher, 20–22, 25–26, 173 n9, 174 n19 High Performance Systems, 96 Hindelang, Michael J., 93 Hirschi, Travis, 34–35, 77 Hoffman, Peter B., 36, 84–85
223
Ignatieff, Michael, 24, 30 Innes, Christopher A., 36 Irwin, John, 5, 75, 97, 157, 161, 172 n8 Jackson, George, 53 Jackson, Margaret Windsor, 182 n10 Jackson, Pamela Irving, 75 Jacobsen, Chanoch, 12, 98, 171 n6 Jacoby, Joseph E., 35, 79, 82, 89–90 Johnson, Philip D., 39, 46–47, 49 Jones, Bill, 41–43, 47, 55, 177 n4 Jones, David A., 13, 27 Kadish, Sanford H., 31, 39, 48 Kaae, Jennifer, 52 Kamin, Sam, 5 Kane, Thomas R., 85 Kannensohn, Michael, 51 Kempinen, Cynthia, 182 n1 Kirschheimer, Otto, 26 Kittrie, Nicholas N., 30 Kleck, Gary, 97 Koch, Gary C., 10 Kovandzic, Tomislav V., 131 Kozol, H. L., 78 Kuhn,Thomas S., 4, 17–18, 25, 30 Kuntz, William Francis (II), 21, 173 n10, 175 n21 Kuzmics, Helmut, 24, 174 n16 Lamott, Kenneth, 44, 178 n8 Langan, Patrick A., 50, 93 Langbein, John H., 21, 25, 173 n11, 176 n28 Larson, Richard, 10, 95 Laub, John 75 Lea, John, 14–15 Legislative Analyst’s Office (California), 10, 57, 132, 178 n7 Leonard, Elizabeth Dermody, 49, 185 n10 Lesser, Selma, 51 Lewis, J. L., 26 Lipton,D., 18, 30 Liska, Allen E., 75
224
Author Index
Litvan, Laura M., 57, 88 Lobingier, Charles Sumner, 20 Lotke, Eric, 43 Lowman, John, 15 Lowrance,William W., 74–76 Maguire, Kathleen, 5, 78, 88, 92 Marquart, James M., 83–84, 182 n11 Martinson, Robert, 18, 30, 175 n24 Mathiesen, Thomas, 18, 171 n2 Matthews, Roger, 14–15 McCorkle, Richard C., 156 McGowen, Randall, 19, 25–26, 28, 173 n12 McIntyre, Jennie, 76, 94 McIntyre, Jennifer, 5 McKelvey, Blake, 21, 28, 175n McGee, Richard A., 47–48, 171 n4 McWilliams, Carey, 44 Megargee, Edwin I., 80 Menzies, Robert J., 90 Merlo, Alida V., 62 Merton, Robert K., 14, 176 n33 Messinger, Sheldon L., 39, 45–47, 49 Mihata, Kevin, 97–98 Miller, JoAnn L., 77 Miller, Marc, 63, 66–67, 74, 77, 181 n2 Miller, Stuart J., 29, 80 Minor-Harper, Stephanie, 92 Mitford, Jessica, 30 Monahan, John, 35, 64–65, 75, 79–82, 89–90 Moore, Mark H., 18, 67, 76–77, 80, 182 n7 Mordt, Gabrielle, 98 Morris, Norval, 7–8, 30, 39, 63, 66–67, 74, 76–77, 171 n2, 173 n8, 181 n2 Myers, Martha A., 75 Nagin, Daniel, 32–33, 35, 49, 62, 68, 78, 176 n30 National Council on Crime and Delinquency (NCCD), 76, 78 National Reformatory Union, 25 National Research Council, 78 Neal, David, 39, 50 Neiman, Max, 85–86
Ohlin, Lloyd E., 10, 95 O’Malley, Pat, 7, 36 Owen, Barbara A., 44, 92, 123 Packer, Herbert L., 18, 61, 66, 171 n2, 173 n6 Palmer,Ted, 6, 176 n35 Pannell, William, 49 Pastore, Ann L., 5, 78, 88, 92 Patrick, Steven, 95, 97–98 Pearce, Frank, 75 Petee, Thomas, 161 Peters, Edward M., 20–21, 25 Peters, Michael H., 95 Petersilia, Joan, 36, 62–63, 66, 145 Pens, Dan, 43 Philips, David, 24, 174 n17 Pillsbury, Samuel H., 156 Pisciotta, Alexander W., 175 n22 Pollard,Vic, 55 Pratt, John, 74–79 Pressman, Steven, 52 Prince,Walter Franklin, 22 Proctor, Jon L., 85 Quetelet, Lambert Adolphe Jacques, 27, 174–175 n20 Quinney, Richard, 14 Rafter,Nicole Hahn, 78–79 Rawls, James J., 44 Reichman, Nancy, 7, 36 Remington, Frank J., 10, 95 Rennie, Ysabel, 74–76, 78–80 Resetar, Susan, 171 n6 Rigakos, George S., 36 Riker, Alissa, 5 Robison, James O., 83 Roemer, John, 56 Rossi, Peter H., 36, 62–63, 69, 77, 161, 179–180 n3 Roth, Martin, 53 Rothman, David J., 7, 21–22, 24, 26–32, 48, 173 n8, 174 n18 n19, 175 n21 n27 Rotman, Edgardo, 27–28, 175 n21, 176 n35 Rusche, Georg, 26
Author Index Russell, Bertrand, 154 Sampson, Robert J., 75 Sarbin, Theodore R., 74 Schelling, Thomas C., 97, 172 n7 Schiraldi,Vincent, 5, 43, 95, 136, 183 n11 Schrag, Peter, 40–41, 52, 57–58, 177 n6, 179 n21 Schrager, Laura Shill, 181 n5 Scott, P. D., 74, 76, 90 Scott, S. P., 20, 173 n7 Sellin, Thorsten, 14, 19–21, 35, 62–63, 69, 161, 173 n7 Shah, Saleem A., 65, 76–77, 80, 180 n5 Shannon, Lyle W., 36, 62, 69 Shaw, A.G.L., 22, 25 Shelden, Randall G., 66 Sherman, Michael, 6–7, 13 Shichor, David, 176 n35 Shinnar, Reuel, 181 n10 Shinnar, Shlomo, 181 n10 Short, James F., Jr., 181 n5 Simon, Jonathan, 3, 7, 36 Simpson, Jon E., 77 Sleffel, Linda, 80 Smith, Douglas A., 93 Smith, Gerald W., 83 Sorenson, Jonathan R., 84 Spelman,William, 5, 8, 35, 68–70 Spierenburg, Pieter, 21–22, 26, 172 n1, 173 n12 Stafford, Mark, 94 Starr,M. B., 44 State Board of Prison Directors (California), 45–46 Steadman, Henry J., 35, 65, 78–79, 81–83, 89–90, 92 Steen, Sara, 172 n8 Steury, Ellen Hochstedler, 161 Stolzenberg, Lisa, 5 Sutherland, Edwin H., 176 n33 Tarde, Gabriel, 34, 176 n33 Taugher, Brian, 32 Teller, Paul, 96–97
225
Thornberry,Terence P., 35, 79, 82, 89–90 Tombs, Steve, 75 Tonry, Michael, 5, 36, 39, 97 Trebach, Arnold S., 34 Turk, Austin T., xi, 14, 75, 79 Turner, Jonathan H., 100 Turner, Ronald E., 10–11, 95 Turner, Susan, 15, 58, 76, 87, 98, 113 Ulmer, Jeffery T., 5 U.S. Department of Justice, 94, 99, 185 n5 van den Haag, Ernest, 30 Van Dine, Stephan, 80 Visher, Christy A., 66–67, 69 Vold, George B., 14 von Hirsch, Andrew, 6, 18, 30–33, 64, 66–67, 171 n2, 180 n6 Walker, Nigel, 6, 18, 74, 171 n2, 173 n6 Warner, Barbara D., 75 Warr, Mark, 94, 183 n6 Weidenhofer, Margaret, 22, 25, 173 n12 Welsh, Jane, 64, 68 Wenk, Ernst. A., 83 Whitlow, Karen L., 85 Whyte, Edward H., 45 Wicharaya, Tamasak, 29, 33 Wildavsky, Aaron B., 155 Wilkins, Leslie T., 76 Wilks, J., 18, 30 Williams, Tricia, 77 Wilson, James Q., 4, 18, 31, 34 Wilson,William Julius, 93 Wines, Enoch C., 27, 29, 44 Wolfgang, Marvin E., 35, 62–63, 69, 77, 94, 161 Woznicki, Robert, 178 n11 Wright, James D., 36, 62–63, 69, 161, 179–180 n3 Young, Jock, 14–15, 176 n36 Young, Warren 74–76, 80 Yu, Jiang, 75
226
Author Index
Zalman, Marvin, 29 Zedlewski, Edwin W., 5, 70–71 Zimbardo, Philip G., 56, 87, 145, 155
Zimring, Franklin E, 5, 7–8, 13, 33–36, 50, 67–68, 70–71, 94, 157, 171 n2 n3, 176 n29, 181 n12